摘要:

THE FREYEZING POINT OF TRIPLE ALLOYS. 65 XI.-The Freezing Point of Triple Alloys. By C. T. HEYCOCK and F. H. NEVILLE. Gold a.nd Cadmium. IN the Chemical Society’s Journal (Trans., 1891, 59, 936), we de- scribed the peculiar fluctuations in the freezing point of mixtures of gold and cadmium dissolved in tin, and we showed that many of the6 6 EEYGOCK AND NEVXLLE: effects could be explained by the assumption that the two metals combine t o form a compound of the formula AuCd. This substance we afterwards obtained in a pure state (Trans., 1892, 61, 914), and we have since found that the beliaviour of these two metals to each other is substantially the same, whether they are dissolved in tin, bismnth, thallium, or lead. From the curves and tables now given, i t will be seen that if we make a solution oE either gold or cadmium in any of the above-men- tioned solrents, and then add doses of the complementary metal, cadmium or gold respectively, the same phenomena repeat themselves.That is to say, the addition of the third metal sometimes causes a small further fall in the freezing point, but, as more of this metal is added, the freezing point begins to rise, and reaches a maximum when an equal number of atomic weights of gold and cadmium have been added. This maximum freezing point is the same whether gold or cadmium is added first, and is independent of the concentration,* so long as we obser-re the condition of maintaining an equal number of atoms of gold and cadmium in the mixture. After the maximum freezing point has been reached, a further addition of either metal causes a fall, but, as we see in the curve with thallium as solvent, the maximum freezing point can be recovered by adding a sufficient aillount of the complementary metal.Whatever be the explanation OE these phenomena, we here see that the reaction between the two metals is remarkably independent of the nature of the solvent. As the new solvents give results so similar to those recorded in o u r previous paper, where tin was the solvent, we feel no doubt that the cause is the same, namely, the formation and partial precipitation of the compound AuCd, and that this sn5stance exists in solution in equilibrium with free gold and cadmium, uhat- ever be the solaent. It is seldom possible ta study the formation of the same compound in so many solvents, and the uniformity of the results appears strongly to bear out the view that solution in itself is a physicai, rather than a chemical, action. Silver and Cadiniurn.The behariour of silver and cadmium, when dissolved together, is recorded in the Tables V-VIII (pp. 71-73) and Figs. 5-6t (p. 68). It will be seen that these two metals produce in the freezing point of the solrent changes very similar to those caused bj- gold and cadmium. In each case, the addition of the t.hird metal causes a rise, which reaches a maximum value and then declines as more of this met'nl is added. * The maximum is not reached in very dilute solutions. t See " Explanation of the Charts," p. 74.TEE FREEZING POINT OF TRIPLE ALLOYS. 67 Although we have not much experimental proof, yet we have no doubt that, as with gold-cadmium, t h i s maximum freezing point will turn out t o be constant for each solvent, independent of the absolute amounts of silver and cadmium present, but reached when the ratio of these two attains a certain value.In tin or lead, this maximum is reached when the ratio of the two metals is exactly 2Ag: Cd. In thallium, the ratio is nearly the same, and we think that experimental errors would account for the discrepancy. I n bismuth, the ratio at the highest point is not far from 4Ag : Cd. The experimental error is probably small in the bismuth series, but, without further trial, we do not feel certain that this ratio would be maintained a t other concentrations. TABLE L-Cadmiitm aid Gold in 400 grams of Lead.Total weight of gold present. 0 Y Y 7 9 7 9 9 , 9 , Y9 0 G85 2 -083 3 '037 5 -089 5 -717 7 -021 9 -353 14 -04.1 18 -70% 21.631 25 -91'4 Total weight of cadmium preaent. Atom of gold per 100 of lead. 0 9, 97 Y ? I 7 99 9 9 0 .$58 0 *546 0.797 1.335 1-500 1 .M2 2 -453 3 -683 4 -906 5 '674 6 -797 Atoms of cadmium per 100 of lead. Freezing point of solution. 327 -60' 327 -24 326 -85 32a -80 323 '86 319 *62 315 '48 308 7ti 310 -04' 312.W 313 -441 314-60 315 -10 316 -06 317 -25 318 -35 318.55 316 *9L 322 -40 * As soon as the first portione of gold were added, a grating of the stirrer was noticed, and other proofs of a precipitate which rapidly increased. Much preci- pitate was found in the crucible at the termination of the experiment. If we compare together the curves for gold-cadmium, and silver- cadmium, in bismuth, and then compare them with all the others, we shall see that these two stand somewhat apart.We do not yet know whether this is due to the specific action of bismuth as a solvent, or to a cause of experimeiital error, which we hace, hitherto, failed to detect. Curve 5 of cadmium added to a saturated solution of silver in tin is singularly like the saturated gold-cadmium curves of onr prerious p p e r : the initial fall caused by the cadmium, followed by what may68 HEYCOCK AND NEVILLE: be a short f l d , is to be seen, also, in the gold-cadmium curves. We haye traced experimentally a 3-atom silver and cadmium cnrve in tin, where, j u s t as in the 3-atom gold-cadmium curves, there is no raaxi- mum freezing point.This curve is not given in the present paper. TABLE 11.-Gold and Cadmium i n 400 grams of Lead. Total weight of gold present. 0 0 '822 0.632 1 -i42 5 -417 9 -451 12 -563 15 231 9 ) 7 9 Y ? Y 7 Y 9 31 Y1 I1 Y Y 9 9 ?? Y 1 Y l 2 9 7) 9 ) Total weight of cadmium present. 0 Y9 Y Y ?f YY Y Y 99 7 9 0 -159 0 '441 0 -730 1 '083 1 -756 3 -345 5 -418 6 -630 7 -6i5 8 -666 11 '841 13.364 16.108 It) *801 21 -463 9 -78% Atoms of gold per 100 of lead. 0 0'085 0 *l66 0 '457 1 -521 2 *479 3.292 4 -0oo > > J Y Y ) Y * 1, I > 9 7 Y Y 9 , 7 ? 9 Y Y 7 9 Y , 1% 7 Y Atoms of cadmium per 100 of lead. 0 9 9 Y Y Y7 9 , Y Y $ 1 1, 0 -073 0 -204 0 -337 0 -5uo 0 -810 1 -643 2 -500 3 -059 3 541 4 -0oo 4 -514 5 -186 6 -165 I ado 8 -675 9 -903 P ..- Freezing point of solution.327 '64' 327 -10 326 '54 324 -TO 318 -45 311 -61 306 -44 301 -84 301 -77 301 -57 302 -53 303 -69 305 -51 309 -62 314 -13 316 -4Q 317 -72 318 '54 318 '78 318 -47 317 -08 315 -46 314 -90 314 -20 A mass of crystalline precipitate was found a t the termination of the experiment. I n Series 6 and 8, silver was again added from the point X, and produced, as we expected a rise. All these curTes contain details that we cannot interpret, but we. are disposed to explain the general character of 5, 6, and 7 by assuming that a sparingly soluble compound, Ag,Cd,* is formed, which call only exist in the presence of free silver and cadmium. Much of what was said in our paper on triple alloys of gold, cadmiurn, and tin (Zoc. cit.) would then be applicable. The only point we will here consider is the position of the summit C.Because the summit C is reached at a concentration of 2Ag: Cd, we are hardly entitled to concliide at once that the compound AaCd * Our attempts to isolate this compound in the same way as SuCd hare so far failed.68 HEYCOCK AND NEVILLE: be a short f l d , is to be seen, also, in the gold-cadmium curves. We haye traced experimentally a 3-atom silver and cadmium cnrve in tin, where, j u s t as in the 3-atom gold-cadmium curves, there is no raaxi- mum freezing point. This curve is not given in the present paper. TABLE 11.-Gold and Cadmium i n 400 grams of Lead. Total weight of gold present. 0 0 '822 0.632 1 -i42 5 -417 9 -451 12 -563 15 231 9 ) 7 9 Y ? Y 7 Y 9 31 Y1 I1 Y Y 9 9 ?? Y 1 Y l 2 9 7) 9 ) Total weight of cadmium present.0 Y9 Y Y ?f YY Y Y 99 7 9 0 -159 0 '441 0 -730 1 '083 1 -756 3 -345 5 -418 6 -630 7 -6i5 8 -666 11 '841 13.364 16.108 It) *801 21 -463 9 -78% Atoms of gold per 100 of lead. 0 0'085 0 *l66 0 '457 1 -521 2 *479 3.292 4 -0oo > > J Y Y ) Y * 1, I > 9 7 Y Y 9 , 7 ? 9 Y Y 7 9 Y , 1% 7 Y Atoms of cadmium per 100 of lead. 0 9 9 Y Y Y7 9 , Y Y $ 1 1, 0 -073 0 -204 0 -337 0 -5uo 0 -810 1 -643 2 -500 3 -059 3 541 4 -0oo 4 -514 5 -186 6 -165 I ado 8 -675 9 -903 P ..- Freezing point of solution. 327 '64' 327 -10 326 '54 324 -TO 318 -45 311 -61 306 -44 301 -84 301 -77 301 -57 302 -53 303 -69 305 -51 309 -62 314 -13 316 -4Q 317 -72 318 '54 318 '78 318 -47 317 -08 315 -46 314 -90 314 -20 A mass of crystalline precipitate was found a t the termination of the experiment.I n Series 6 and 8, silver was again added from the point X, and produced, as we expected a rise. All these curTes contain details that we cannot interpret, but we. are disposed to explain the general character of 5, 6, and 7 by assuming that a sparingly soluble compound, Ag,Cd,* is formed, which call only exist in the presence of free silver and cadmium. Much of what was said in our paper on triple alloys of gold, cadmiurn, and tin (Zoc. cit.) would then be applicable. The only point we will here consider is the position of the summit C. Because the summit C is reached at a concentration of 2Ag: Cd, we are hardly entitled to concliide at once that the compound AaCd * Our attempts to isolate this compound in the same way as SuCd hare so far failed.Jozr rm.C h m . SOC. fib. I894 HEYCOCK & NEVILLE. SILVER AND CADMIUM.THE FREEZING POINT OF TRIPLE ALLOYS. 69 Total weight of bismuth present. 350 365 380 395 410 4 25 Y Y ?I 7 9 Y ? Y ? Y Y 7 7 Y Y Y * 1 7 1 7 Y ? > I JY Y 7 I t l Y Y f Y Y 11 Y 7 Y l Y Y Y Y t t > 9 7 9 9, Y Y Y Y Y Y Y Y Y l ? Y ?I * A has been formed. mately in the following manner. But we can arrive at this conclusion more legiti- ~ Total weight Total weight Of k;,”i”,$ Freezing point of of gold of cadmium present. present. perlOOof per of bismuth. bismuth. -- ----- -- 0 0 0 0 267 a 5 4 O 2 *252 1 9 0 -651 l Y 266 ‘24 4 -826 Y l 1 ’340 9 ) 264 -92 9 -621 9 9 2 ‘569 Y Y 262 ’52 14 -823 l l 3 ’813 I t 260 -03 20 *152 5 ‘001 Y Y 257-88 Y 9 0 ‘227 Y Y 0 ‘10 258 ’39 Y , 0 -4.67 l t 0 ‘2C4 258 -96 Yr 0 -936 19 0 ‘408 260 -00 l t 1 -884 Y l 0.822 261 -67 7 , 2 -917 I ? 1 ‘273 263 -11 Y , 3 -903 l ! 1.704 263 -90 Y 7 4 -913 9 , 2 -145 263 -99 Y, 5 -207 7 7 2 -274 263 ‘98 t l 5 -720 Y Y 2 -498 263 ‘99+ 6 -630 Y l 2 ‘895 264 -16 7 -186 Y l 3 -137 264 *36 97 7 -677 Y Y 3 ‘352 264 -81 > I 8 *079 Y Y 3 -528 265 -01 Y l 8.396 73 3 -666 265 ‘15 Yl 8 -768 7 7 3 -828 265 -39 -360 Y l 4 -087 265 -81 979 -577 Y l 4 -617 266 -30 11 -238 Y Y 11 -801 Y Y I 5.152 266 -29 Y! 12 *332 l .5 ‘385 266 ‘00 t l 13 -199 19 5 -763 265 ‘64 Yt 14 -204 l Y 6 -201 265-10 Y 7 15 *2W l Y 6 -640 264 -30 $1 16 -376 I9 7 -150 263 ‘43 Y Y 17 -628 I 7 -696 262 ‘46 >l 19 -044 Y > 8 -314 261 ‘41 IY 20 -639 Y I 9 -011 260 -35 > Y 22.274 t? 9‘724 259 *Oo I , 23.594 9 ) 10 ‘30 258 ‘05 Yt 25 -622 Y Y 11 -19 256 -66: Y Y 27 -760 Y - 12 -12 255 -20 37 29 -590 Y Y 12 -92 254 -17 l Y 33 -954 Y Y 14 *825 252 -60 Y Y 0 -?la 7 - 0 ‘05 258 -20 l Y Y * Y Y Y ) Y Y 4 ‘357 266 -15 266 ‘36 266 ‘49t Y l Y l Y precipitate here formed.TABLE 111.-Gold and Cadmiwn in Bismuth.70 HEYCOCK AND NEVILLE: TABLE TV.-Citdmiunt and Gold in 6G-34 grams of Thallium. Gei ssler Thermometer. Total weight of gold present. 0 9 ) Y 9 Y 9 Y Y Y Y 0 -%8 0 *408 0 -728 1 '051 1 *409 1 '854 2 -234 2 -514 3 *013 3 -627 7 ) 7Y Y Y Y Y Y , Total weight of cadmium present. 0 0 -104 0 -201 0-517 0 -998 1.328 3 ) Y 7 7Y Y Y 7) Y Y Y Y Y Y Y Y > Y 9 , 1 -999 2 -046 2 -129 2 -229 2 -511 Atoms of gold per 100 of thallium. 0 7 ) Y 9 Y Y Y 7 Y Y 0 "700 1 -250 1'806 2 '419 3'183 3 '835 4 -316 5 -172 6 '226 Y Y Y Y Y 9 9 Y Y Atoms of cadmium per 100 of thallium.0 0 -312 0 -604 1.5657 3 -006 4 -000 6 &3 6 '166 6 '416 6 T I 4 7.568 Freezing point of solution, 301 %7" 300.18 298 '90 294 -86 288 -10 284 '52* 285 '42* 284 .97 287 '28 289 -11 291 -07 293 - 16 294 -68 294 '29 291 -17 285 -69 294 '40 299.53 294 -86 294 -66 293 37 294 '77 * Between these readings, there was an interval of two days, and the block was st,rongly heated during a part of the time. The discrepancy ip probably due t o two causes, a rise in the zero of the thermometer and, perhaps, mme oxidation of the cadmium. Suppose that the compoiind Az,fiCd, is being formed, but that i t can only exist i n the presence of free silver and cadmium.Then, if we have present in solution x atoms of free silver, y at.oms of free cadmium, and z molecules of the compound Ag,,Cd,, the theories of chemical equilibrium accepted at present indicate that 2, y, and z should satisfy the equation xmyn = KZ, where K is a, constant so long as the temperature does not vary much. Let UTJ consider only that part of the curve where the liquid is saturated with the compound ; then, for this part of the curve, z is constant By adding silver or cadmium, we can vary x and y, but only subject to the condition that r%f is constant. Now, a maximum freezing poiiit that is a summit C on the curve corresponds to a minimum number of molecules in solution. That is to say, at C we must have x + y + z a minimurn, or as z is con-THE FREEZE@ POIXT OF TRIPLE ALWYS. 71 stant, we must have a: + y IL minimum snbjeot to the condition x m y * constant. This leads at once to the result that at the summit C the relation between x and 3 is my = ?Ax.Now, if a and c are the numbers of atoms OE silver and cadmium that have been added to the mixture, and P the number of molecules o€ silver-cadmium that has been formed, then a = 7nP t x, c = 71P + y, and maltiplying by n and m and subtracting, we see that n.a = mc a : c = m : n. TABLE V.-f&lver and Cadmium in 300 grpm of Tin. Total weight of eilver present; 0 5 -478 10 -956 13 -695 Y2 Y Y? 2 2 Y 2 ) 7 2 Y Y 22 7 7 7 9 9 1 7 Y 2 > > ;> Y ? Y Y 3 7 7 7 7 7 Y Y Y 3 2 Y 7 ? f 7 2 Total weight of cadmium present. 0 Y Y 39 0 -;85 0 '83.4 1 -424 1 -993 2 -.562 3 -132 3 -701 4 '270 4 -a40 5 -970 7 -118 8 '256 9 -395 10 -534 11-673 13.950 16 -223 19 -075 21 -922 24 -769 25 -908 26 -4'77 27.901 29 -824 30 -748 32 -595 36 -442 Atoms of silver per 100 of bin.- 0 2 4 5 9 1 > Y Y Y > I Y2 Y 7 2 7 7 I Y 2> 7 7 7 7 7, 2 ) 7 9 YY 7 3 ,, 3 9 > 7 x2 27 > 2 9 2 9 9 2 > Y Atoms of cadmium per 100 of tin. 0 J 9 3 ) dfl 0 -3 0 -5 0 -7 0 -9 1 -1 1 -3 1 -5 1 -7 2 -1 2 -5 2 -9 3 -3 3 -7 4 -1 4.9 5 -7 6 -7 7 9 8 '7 9 -1 9 '3 9 -8 10 -8 10 -8 11 B 32 B Freezing point of solution. 23 I -49" 225 '95 221 '12 221 *04 220 -84 220 -68 220 -33 219 -57 219 '56 219 -64 219 -73 219 -86 219 '84 21 9 '93 219 '98 2L9-97 219.93 219 -86 218 77 219 -51 219 '16 218.88 21 8 2 5 218 -00 W 8 '22 218 -12 217 8 9 217 -62 217 -35 216 $7 216 Q272 HEYCOCE AND NEVILLE: Therefore the snmmit will always be reached when the quantities of the two metals added are iu the ratio given by the formula of the compound formed.In the case of silver and cadmium, me find that in t,he solvents tin, thallium, and lead the body formed is AglCd, whilst in bismuth there is, perhaps, the body AgdCd. In the main, me see that the reaction between silver and cadmium is, like that of gold and cadmium, independent of the nature of the solvent. TABLE VI.-Silz.er in 61.18 gmms of Thallium, and then Cadmium added.* Total weight of silver present. --- 0 1 -326 7 7 Y! 77 77 7 1 1 ) Y I 7 7 1 sib3 1 598 1 -681 1 939 1 -821 1 -889 1 -967 2 -060 2 -165 4 -165 Total weight of cadmium present. 0 0 -272 0 424 0 -515 0 569 0 '602 0 -662 0 5'02 0 -969 7 9 17 I Y Y1 7 ) 7 ? 7 7 3 Y l 1 ) Atoms of silver per 100 of thallium.0 4.1 7 9 77 9 9 Y9 Y, 7 7 7 1 9 9 4 &5 4 '9.10 5 '196 5 -375 5 '628 5 -838 6 '080 6 '367 6 '692 12 '875 Atoms of cadmium per LOO of thallium. 0 0 k 7 0 -810 1-263 1 -534 1 '696 1'793 1 -971 2 -09 2.886 Y Y 7 9 9 1 9 1 17 7 7 11 1 9 11 91 Freezing point of solution. 301 -93' 288 *lot 290.19 292 -04 293 -10 293 -2p 293 -27 293 '28 293 -18 293 -08 291 -86 292 -29 292 -4.9 292 *81 292 *76 292 '86 292 '97 292 -99 293 -04, 293 -06 293 '09 * This series, hke all the thallium experiments, was canied out on too smaU a scale for accurate results. f More than saturated. TABLE VII.-Ssa'lver and Cadmium in 400 grams of Lead. Total weight of silver prewnt. - 0 0 -167 0 -687 1 -d91 3 -907 7 -149 Total weight of cadmium present. Atoms of silTer per 100 of lead.Atoms oh cadmium per 100 of lead. 0 0 '08 0 '329 0 '906 1 '872 3 ' a 5 0 1 ) 9 ) 7 9 1 8 7J Freezing point of solution. ~- 327 *so" 327 -18 325 -62 322 -18 316 TO 308 *64THE PREEZlSG POINT OF' TRIPLE ALLOTS. 73 ~ Total weight of silver present. Total weight of cadmium present. 0 - 0 -a62 1 -404 1 -934 2 -4% 3 -502 4 '576 6 -275 f -771 I -951. 8 -492 9 -470 11 -603 14 -688 18 *634 20 "772 24 - 9 i i Atonis of silver per 103 of lead. .5 a 2 (i -002 --- 7 7 > 7 7 Y 7 P 7 1 7 7 7, 7 . 7 - > 7 7 &onis o€ cadmium per 100 OF lead. 0 o .%8 0 -648 0 -892 1.14'7 1 -61-5 2 -112 2 9395 3 -12.5 3 -486 3 *918 4 -370 5 -354 k; -767 8 -395 9 *536 11 '92.5 F r e I? z i n ,D point of solutio 11.TABLE VIII.-Silcer and Cudiiziunz in L'iism u f l ~ Thermometer, Hicks' 12 A. Total weiglit of bismuth present. Total weight of silwr present. 0 3 '838 7 -676 10 -635 -- 9 1 7 J l 97 7 ) J 9 7 9 7 7 '7 7 > r 77 7 Y Y 7 7 77 7 7 14 .k3l Tot a1 weiglit OE cadmium present. 0 --- Y l 7. 0 *553 1 -105 1 -668 2 -210 9 .T63 3 '315 3 *86i 4 -320 4 -9i;j 5 -5:s 6.078 7.193 8 -230 9 -332 3 0 -497 11 -6 13 13 -812 16 -023 20 -443 24 363 9 . 303 -Mi" 303'15 306 -62 308 67 310.32 311 '66 313.55 31 a -99 31.5 -75 315 *7!J 315 '70 315'39 314 98 313 48 311.99 311.13 311 '33 310 ,311 &oms of cadmium bismuth. pel- 100 of Freezing point oi solution. VOL. LXV. G74 HBTCOCK AND XEVILLE: EXPLANATION OF THE CHARTS. The curves gire the freezing points of soliltions of gold nit11 cadmiu~ii.ancl of silver with cadmium in different solvents. Each black dot records the concentration and freezing point of a yarticulal. mixture. In eT-ery case, 100 atomic weights of solvent arc assumed to be present. The horizontal distance of a point measured from the left records the total number of atomic weights of foreign metals, other than the solrent present, each square corresponding to one atomic weight. Tho vertical distance of a point from the ler-el of the first point on the curve records the fall or rise in the freezing point, each square being one 1" centigrade. Along the t h n iine, cadmium is being added : along the thick line, gold or silver- I n Curve 4, we start with a mixture of 4 atoms of cadmium and 100 atoms of thallium; in other cases the curve begins witn the pure solrent.The point A marks where we begin to add the third metal. C is the summit of the curve. The tables are numbered to correspond with the curves. I n the tabIes all weights ore in grams, and temperatures in degrees centigrade. The point X mnrIis where we begin again ta add the second metal. -1lunzinizcnt and Gold dissolced iu Tin. I n striking contrast to the foregoing curves, which owe their character to the partial dissociation of the compound fonned, we non- give curves showing the freezing points of alloys made by dissolving aluminium in tin and then adding gold. Fig. 9 commences by the addition of aluminium to tin. We see that the aluminium produces an atomic fall of 1*4", which is half the normal fall produced by other metals in t'his solvent.This, as we suggested in oiir previous paper (Trans., 1890, 57, 392), may be explained by assuming that the met.al is present, in diatomic molecules as AI,. The addition of aluminium was elided a t A, almost exactly at the point of saturation of the tin. Gold was then added, as recorded by the thick line. It at once produced R rise, the maximum freezing point C, which was identical with that of pure tiiz, being reached with 2 atoms of aluminium and 0.9 atom of gold, or, in other words with a concentration of AI,Au, 9. A further addition of gold produced a fall, which, reckoned from the highest point, is exactly that of gold in pure tin. In Fig. 10, a total quantity of 4 atoms of aluminium mas added up to the point A, saturation occurring at 2 atoms.The first por- tions of gold produced no effect on the freezing point of the mixture until 0.95 atom had been added, to the point B, but on adding more there was a rapid rise ; the highest point, again identical with the freezing point of pure tin, being reached with 1.9 atoms of gold. When we add more gold from this point C, we seem to be starting afresh with a solution of pure tin, and we get the atomic fall 2-86", which is identical with our old numbers. The obvious explanation of these resnlts is, that the gold and0 c3 x! 0 0 0 > W Ir - TEE FREEZIKG POINT OF TRIPLE ALLOTS. 1 .3 Atoms of aluminium per 100 of tin. --- 0 0 . 5 1 -5 0 -2 ? > > 7 7 Y 7 2 7 '3 7 7 aluminium form a stable insoluble compound of the formula Anal,, the gold completely removing the aluminium from solution.There is no curved part to the louiis of the freezing points, inasmuch as free gold and aluminium cannot exist together in solution. That rather less gold is needed to reach C in both figures, or B in Fig. 10, than the amount corresponding to the compound AuAlz, may fairly be put down to the fact that all the loss by oxidation in making the alloys, and probably all the original impurity, falls on the aluminium. The fact that in Fig. 10 the gold from A. to B produced no effect agrees very well with our view that the aluminium flat up to A. is due to saturation of the tin with aluminium. If this be so, then, although the first quantities of gold up to B precipitate aluminium, yet some of the excess mill come into solution and the freezing point will not alter.Bat by the time 1 atom of gold has been added, the excess of aluminium rill have been used up to form the body Al,Sn, and further addition of gold must decrease the aluminium in solution, and so produce a rise in the freezing point. The precipitate AuAl, is, no doubt, the purple alloy discovered and examined by Professor Roberts-Austen. He finds for his alloy the formula A12A~0.95. Our values deduced from the freezing point curves are :- At C, Fig. 9 . . ............... AI,An,.9 ,, ,, 10 ................. L4.12Au,,.L,5 A t B, ,, 10 ................. A12duo.95 TABLE TX.-Alumiwium ctrzd Gold in Tin. Thermometer, Hicks' No. 12. Total weight of tin present. -- 300 3 10 320 330 > Y J > f , 7 I ? 7 f > > Totsl weight of aluminium present -- 0 0-335 1.101 1 -6u5 > 9 7 7 1 . > 2 ) 7 7 ) Total weight present. of gold Atoms of per 100 of tin. goia Freezing point of solution. -- 231 -55" 230 -90 229 4 8 228 67 228 -97 229 -29 229 -90 230 -55 231 - 20 231 -55 231 -34 230 *76 229 -6176 HOOKER AXD CARNELL : THE COXDENSATIOK Tables IX and X give the experimental numbers from which the Total weight of tin present. 250 275 280 300 J 7 9 9 97 Y 9 f $ 9 7 9 > ) Y Y 9 9 alumininm curves were drawn. Total weiglit of aluminium present. 0 1 261 1 -605 2 .is2 7 7 > 1 9 9 Y Y 7 7 Y 9 Y ? 9 2 7 7 9 TABLE X.-Aluminium awd Gold in Tiu. Thermometer, Hicks' No. 11. Total weight of present. gold -- 0 Y Y Y 2 *go6 3 -750 4 260 4'761 5.263 7 -769 9 -022 9 523 10 '024 12 -530 Atoms of aluminium per 100 of tin. -- 0 9 -0 2 ' 5 4 -0 Y ) 9 7 Y i 7 7 7 7 Y ? ? Y 3 , I Atoms of per 3 0 0 of tin. gold 0 3 7 >7 dY5 0 -75 0 -85 0 -95 1 -05 1-55 1 -8 1-9 2 -0 2 -5 Freezing point of solution. -- 231 -65' 228 '72 228 -69 228 -70 228 *'73 228 -7 1 228 '73 228 -74 229 - 07 230 5 7 231 -36 231 -59 231 -29 229 -86 Sidney College, Cambridge.

ISSN:0368-1645    

DOI:10.1039/CT8946500065

出版商:RSC    

年代:1894

数据来源: RSC

摘要:

76 HOOKER AXD CARNELL : THE COXDENSATIOK XII.-Contwviou of Ortho- iiiti, Pnru-, c w c l of Purci- deusa t i o n of A ldeh y dcs 10 it IL p- Hyt7 I 'oxy- a- imph t Ii CI - pumone. ilzto Ortho-quinollc' Dc~*it~~iti~cs." I. The? COSZ- By SAMUEL C. HOOKEI; and WILLIAM C. CARXELL. Is the course of researches in the lapachol group, already partially communicated by one of us to the Society, nunierous compounds of the general formula 0 0 have been isolated, from which dehFdratiiig agents, preEerably dihtte or concentrated mineral acids, rery readilj- remore wntey or hydrogen * Compare Tram., 1893, 63, foot note p. 431; also Proc., 1893, 13.OF ALDEHPDES WITH B-HYDROXT-a-NAPHTHAQUIXONE. 7 7 chloride or bromide, quantitativeIy converting them, without exception, into internal anhydrides.It was at first supposed tbat the change was a simple one, involving nothing further than the elimination of water; but it was subse- quently found that several of the lapachols, namely, chlorhydrolsp- achol, hydroxyhydrolapachol, and bromhydroxyhydrolapachol, each ga-re two isomeric anhydrides, which proved to be related as follows. 0 Corresponding or a-aniijclride. Pseudo- or 8-anlqdride. In spite of the very varied conditions under which they have been exposed in contact with mineral acids, the remaining lapachols have given in each case only one anhydride, and these have all proved to be P-nzpht,haquinone derivatires. It can, therefore, be generally stated, at least so far as the com- pounds at present known in the lapachol group are concerned, that substances of the general formula I are more readily converted by the action of mineral acids into the p- than into the a-anhydrides, a change which involves the conversion of the para- into the ortho- quinone group.In order to ascertain how far this formation of the &anhydrides would prove general, we have investigated the beha viour of compounds prepared synthetically from P-hydroxy-2-napb tha- quinone. Zincke and Thelen (Ber., 21,2203) first obaeryed that benzaldehydc and /?-h y drox y -a-nap ht haq uinone interact readily, and correctly attributed to the resulting compound the formula- 0 I CbH5 Schoch", one of Zincke's pupils, snbsequen tIy extended the study of this reaction to other aldehydes, and prepared a number of com- pounds having the general formula * " Ueber die Einwirkung Ton Aldehyden und Ketonen auf Oxynaphtochinon," We are indebted to Professor Zincke for Inaugural Disscrtat'm, Marburg, 1858.a copy of this pamphlet.7s HOOKER AXD CARNELL : THE COXDENSATION 0 0 I x In preparing some of these compounds, small quantities of an- hFdridca *are obtained as secondarj- products, t o which the general formula 0 0 0 A/\/\\/V\ I I I I I I v \o/\L&\d'\/ I s was assigned. These anhydrides xwe also in some cases directly ob- tained from the corresponding hycli*oxyl compounds by the action of dehydrating agents ; b u t in every instance Schoch records that the yield as veq- small. The behaviour of the lapachols in forming, as a rule, /%anhydrides in preference to a-anhydrides, rendered questionable the structure attributed by Zincke and his pupils to the anhFdrides obtained by them ; and our experiiiients have conclusively prox-ed that these compounds, or at least all that we haTe examined, are, in realitF, as we anticipated, deriT-atiTes of /%naphthaquinone, to which one or other of the following formula: must be attributed. I * s s c&Aiihydridc.~ ~ - A n h ~ - d r i c l e . The anh-plyides may be prepared quantitatively, or almost so, by the methods successfullF used in t h e lapachol group. In no instance, however, in spite of many experiments. have we been successful in obtaining more than one anhydride from each compound. All the anhjdrides studied condense readily with orthotolu_vlene- diamine, giving characteristic azines, an action which, in this case, can only be explained by the presence of the orthoquinoiie group.Up t o the present time, the azines obtained have resulted from the action of one molecule or the diamine on one of the anhydride, so that, of the abo3-e €ormula;., the first, which is that of an ~/5-anhydi=ideOF ALDEHYDES WITH ~-HYDROXY-C~-NAPHTEAQVINOXE. 79 still containing one pal-aquiiione group, is perhaps the more probable, although we do not regard the second as definitely excluded. As a further proof of the presence of the orthoquinone group, it may be urged that there is a marked difference in the colour of the hydroxyl compounds, and of their corresponding anhydrides. The former, like a-naphthaquinone, are yellow ; the latter, like &naphtha- quinone, are orange or ormzge-red.A very considerable number of compounds of the lapachol group hare been studied, and this change in colonr from yellow to orange or red has proved an unfailing indication of a simultaneous change from the para- to the ortho- c yuinone. It has been shown in the lapachol group, that both the a- and &anhydrides take up water when boiled with dilute alkalis, yielding the same hydroxylapachol--a change, which, in the case of the /%anhydrides, involves the reconversion of the ort ho- into the para- qui none. 0 bv SaOH 0 0. by SaOH Precisely the same change occurs in the case of the synthesised @-anhydrides described in this paper only; it is more diEcult to demonstrate, as the hydroxSl compounds examined very rapidly undergo change in contact even with very dilute caustic soda, a reaction which appears full of interest and will be studied in detail by one of us.0 0 /\/\OH J J o / \ A B e n z y l i d e n e d i h ? ! d r o ~ 1 . y l t a p ~ ~ ~ ~ a ~ ~ ~ i ~ i ~ ~ z e 7 1 I I 1 I v y \L.--'\o/v C,H, This compound was first obtained by Zincke and Thelen (Ber., 21 2293), and subsequentiy more fully examined by Schoch (Diss., Marburg, 1888, p. 11). I n preparing the cornpound for our experi- ments, the directions of Zincke and Theleii were only partially followed. 8 grams of benzaldehyde, 8 grams of hydroxynaphtha- quinone, and 24 C.C. of alcchol were heated in a corked flask on a -.wter bath for about 45 minutes." The heating was then discon- tinued, and the solution diluted with 160 C.C. of alcohol ; the compound * Zincke and Thelen, also Schoch, recommend heating for sereral hours; this is unnecessary, as the action is complete in 45 minutes, and if the heating is extended o-i-ei' a longer period, a portion of the compound is conrerted into its anhydride.80 HOOKER AND CARNELL : THE COXDENSATION shortly aiterwards commenced to separate in minute, yellow tuEts of iriicroscopic needles, vliich, after some hours, mere collected and washed with alcohol.A microscopic examination revealed t h e presence, i n small qaantity, oE orange-red needles,* wliich usually nccom,pany to a somewhat greater extent the additionul substance obtained on concentrating the mnther liquor. The red needles are left for the most pai=t undissolved when the crude substance is re- cr~stallised from alcohol, but it is necessary to repeat the operation, sometimes more than once, in order that t'he separation may be rendered complete. We found the melting poiut of benzylidenedihydroxynaphthaquin- one to be about 230" when rapidly heated, but as the substance decomposes when heated for a short t,irne at temperstures slightly below this, its point of fusiou is influenced by variations in the rate of heating.Benzylidenedihydroxynaphthaquinone gives two distinct sodium derivatives, according as the hydrogen of one or both hydroxyl groups is displaced. The monosodium derivatire cryetallises in orange needles, and may be readily obtained by cautiously adding rz 1 per cent. aqueous solution of sodium hydroxide to a boiling and nearly saturated alcoholic solution of the compound.The addition of the alkali is stopped as soon as the colour of the solution, which is a t first orange, becomes permanently slightly carmine, and rz precipitate of tlie carmine or normal salt begins t o form. Small quantities of benzg 1 idenedih y drox ynap h t ha qui none are then gradually added t o the boiling solution until the carmine precipitate redissolves. 011 cooling, the solution first becomes turbid, arid shortly aftern-nrds orange needles of the sodium compound form throughout its entiye volume. The normal salt, previously obtained by Zincke ant1 Thelen, is n dark carmine-red, crystalline powder. When seen under the microscope, each grain appears as an almost perFectlj- formed octahedron. The very slight solubility of the salt, together with its very characteristic appearance and crystalline form, render it possible to identify minute quantities of benzylideDedihydroxy- naphthaquinone with ease and certainty.It is simply necessary to dissolve the compound in a few drops of hot dcobol, arid add a drop or so of dilute aqueous solution of sodium hydroxide ; if the alcoholic solution be a concentrated one, an abundant precipitate will be almost * This substance, which appeals to Esve escaped the observation of Zinckc, Thelen, and Schoch, differs essentially from the anhydride to be presently described : it dissolves in concentrat,ed sulphuric acid, forming an intenkely bluish-green solu- tion, whereas the anhydride under the same conditions gives a reddish-brow1 solution. We made many unsuccemful attempts to obtain the compound in larger quantities.OF ALDEHYDES WITH ~-HPDROXY-C~-NAPHTHAQUIXONE.81 instantly formed, which should then be examined with a high power. in order that its crystalline form may be recognised. ap-Anhydride of BenzyliLEenedihydl.o~~nap1~ tliaquinone .--T his corn- pound is formed in small quantity in preparing benzylidenedi hydroxg-- naphthaquinone if the time of digestion recommeuded above be much prolonged; i t was in tliis way that Zincke and Thelen obtained it (Ber., 21, 2203); Schoch found that it was produced wheu an alco- holic solution of the hydroxyl compound was heated with a little sulphuric acid ; but even in this case Schoch states that only small quantities were formed (Diss., Marburg. 1888, p. 14). We have found that the compound can be prepared practically quantitative1~- in the following manner :-4 grams of benzylidenedi- hydrosynaphtliaquinone are dissohed in 50 C.C. of boiling acetic acid, to which 2.5 C.C.of concentrated hjdrochloric acid are then verF gradually added ; on continuing the boiling for a few minutes, a heavy precipitate of small, orange-red crystals rapidly forms, and may be collected when the solution has cooled. Wheu slightly washed with acetic acid, i t is obtained in a condition pure enough for all ordinary purposes. A microscopic examination of t h e crystals revealed de- tached, well-formed, acicular prisms. The compound was purified foi- analysis by recrjsfallisation from acetic acid. 0.1837 gave 0.5203 CO, and0.583 H,O. C = 77.24; H = 3 . 3 . C,,H,,O, requires C = 77-52 ; H = 3-35 per cent.When heated in a capillary tube, it commences to darken a t about 245". It dissolves with great difficulty in the ordinary solvents, being almost insoluble in ether, and only very slightly soluble in alcohol, rather more so in chloroform, beiizene, and acetic acid. The anhydride is reconverted into beiizylideiiedihydroxpaphtha- quinone by boiling alkalis. small quantity was moistened with aqueous 1 per cent. sodium hydroxide, thoroughly ground into a, paste ; more of the alkaline solution was then added, and the whole boiled for a few minutes ; after filtration from the unattacked sub- stance, the carmine colcured solution was acidified with acetic acid, and the precipitate collected, and identified, by its characteristic sodium salt, in the manner above described.The anhydride and OrthotoIuylenedianri~~e interact very readily, a thoroughly typical azine being formed : 2 grams of the anhydride, 1 gram of orthotoluylenediamine hydrochloride, and 3 grams of crys- tallised sodium acetate are heated in 70 C.C. of acetic acid. Soon after the acid commences to boil, and before the anhjdride has entirely passed into solution, the azine cgmmences to separate in small clusters of microscopic needles. The conversion is complete i n a few minutes, but, before the boiling is finally discontinued, it is well to82 HOOKER AND CAMELL: THE CONDENSATION ascertain, with the help of the microscope, that no unchanged crystals of the anhydride are present. The substance is then collected, washed with acetic acid or alcohol, and finally with water.For analysis, the azine was twice crystallised from benzene, in which, as in other ordinary organic solvents, it dissolves with considerable difficulty. It was thus obtained in orange-yellow, microscopic needles which can be heated to 245O without undergoing change. 0.1701 gave 8.5 C.C. moist nitrogen at 30° and 764 mm. 0.2189 ,, 0.6516 CO, and 0.0815 H,O. C = 81.17 ; H = 4.13. C,H,,N,O, requires C = 80.95 ; H = 3-96; N = 5.55 per cent. The azine dissolres in concentrated sulphuric acid yielding a, brownish-violet coloured solution, from which an orange coloured salt is precipitated on slight dilution. When zoistened with concentrated hydrochloric acid, it. becomes more deeply orange ; and d e n heated with zinc-dust, and alcohol acidified with hydrochloric acid, a violet coloured solution is obtained, which readily absorbs oxygen, again becoming orange.The azine mas heated with orthotoluylenediamine, under varied conditions, but as no int’eraction occnrred, we were unable to demon- strate the presence of a second orthoquinone group. N = 5-43. 0 0 Schoch has already studied the action of acetaldehyde on hydroxy- naphthaquinone (Diss., Marburg, 1888, p. 24), and obtained a product fusing at 154-157”, which refused to crTstallise, and on analjEis gave figures agreeing with the above formula.. In spite of the analytical figures, the compound must have been in a very impure condition, as ethjlidenedihydroxynaph thaquinone crystallises readily, and fuses at about 190” when rapidly heated ; its fusing point is not? however, quite constant.After a number of preliminary experiments, we have siicceeded in satisfactorily preparing the compound in the following manner :- 5 grams of pulverised hydroxynaphthaq~zinone were heated under pres- sure in a steam bath (a5 about 100°) for five hours with 50 C.C. of alco- hol and 50 C.C. of Schuchardt’s “coucentrated aldehyde” (not absolute). After the heating had been in pi-ogress a short time, the bottle was slightly agitated, so as to ensure the complete dissolution of the hjdroxynaphthaquinone. The new compound separated, as the liquid cooled, in a comparatirely pure condition, in small, heavy crystals ;OF ALDEHYDES WITH ,@-HFDROXT-~-SAYHTHAQUINONE. $3 3.3 grams were obtained. The mother liquor, on evaporation, yielded an additional quantity of the substance, but the condition of this was less satisfactory, and, in order to purify it, it was dissolved as com- pletely as possible in n cold, aqueous 1 per cent.solution of sodium hydroxide. After filt~ation from the insoluble residue,* the sub- stance was reprecipitated without unnecessary delajf by adding ncetio acid, and was tben collected, dried, and crystallised %rom alcohol. E thylidenedihydroxynaphthaquinone can be readily distinguished from hydroxynaphthaquinone by the carmine-red colour of its solu- tion in dilute sodium hydroxide, which presents a sirong contrast to the orange-red solution of the latter. The new compound was recrystallised from alcohol for analysis. It was obtained in irreiular, heavy, golden-yellow crystals which melt when rapidly heated, but not perfeclly sharply, at about 190".The melting point is influenced to some extent by the manner of heating, aiid the fused compound gradually darkens and decomposes even as the temperature falls. 9.1920 gave 0.4950 C02 and 0.0655 H,O. C = 70.31; H = 3.79. C,,Hl,Oe requires C = 70.58; H = 3.74 per cent. up-Anhydride of EthyliderLedihyclroxynaphthnqziinoIae. E thylidenedihydroxynaphthaquinone is readily and almost quanti- tatively converted into its pseudo-anhydride by the action of con- centrated sulphuric acid. 8-5 grams of the compound were dissolved in 15 C.C. of concentrated sulphuric acid, and, after standing about 10 minutes, the solution was poured into a relatively large quantity of water.The orange precipitate was collected on a filter, and as soon as the acid had been mostly displaced by water, it was washed with a 1 per cent. solution of sodium hydroxide to remove any un- changed ethylidenedihydroxynayhthaquinone, and then again with water. When dry, the compound was crystallised from acetic acid; it separated in orange needles, which lost some of their brilliancy of colour when exposed to diff uscd daylight. Analysis gnve the following figures. 0.1881 gave 0,5089 C02 and 0.0606 H20. C = 73-78 ; H = 3.57. C2HI2O,$ requires C = 74.15 ; H = 3.37 per cent. * This consists of a dark red compound which dissolves in concentrated sulphuric ixcid, affording a green solution, and probably corresponcls with the red substance (compare footnote, p.80) observed in the prepnration of bcnzjlidenediliy droxy- naphthquinone, which also g i x s a green solution when dissolved in sulphuric acid. t Ethjliclenedihjdroxynaphthaquinone gradually undergoes change eveu in cold dilute alkaline solution. j: By heating paraldehyde with hydroxynaplithaquinone, Schoch obtained an on analysis approximate to' those required by thc formula C22H1sO; (Diss., Mar-a4 HOOKER AND CARNELL : THE CONDENSATION When boiled with a diInte aqneoas solution o€ sodium hydroxide, the anhydride gradually passes into solution, being reconverted into e t h ylidenedih ydrox y nap hthaqui none, which then rapidly undergoes change. The anhydride interacts rery readily with orthotoluylenediamine ; the resulting compound was obtained in essentially the same manner as the corresponding azine from the /3-anhydride of benzylidencdi- hydroxynaphthaquinone.Two preparations mere made for am- lysis : the one ( a ) was crystallised from benzol ; the other ( b ) from chloroform. ( a ) 0.1276 gave 6.9 C.C. moist nitrogen a t 27" and 762 mm. N = 5-95!. (a) 0.1705 ,, 0.4'317 GO, and 0.0728 H,O. C = 78.65; H = 4-74. (a) 0.1636 ,, 0.4744 ,, ,, 0.0669 ,, C = 79.08; H = 4.54. ( b ) 0.1652 ,, 0.4i43 ,, ,, 0.0675 ,, C = 78 90; H = 4.51. C2,Hl,X20, requires C = 78.73 ; H = 4.07 ; N = 6-33 per cent. The compound crjstallises in minute, orange, woolly needles which dissolve in concentrated sulphuric acid yielding a brownish-violet. coloured solution, and in this, as in other respects, precisely re- sembles the azine already described from the ,%anhydride of benzyl- idenedihydroxynaph thaqiiinone. ( b ) 0.2216 ,) 11.7 ,, $ 7 24.5 ,, 770 ,, N = 5.9s.a 8- Anhydride of Amy lidened ihy droxy nnpht ha q uinone. A mixture of 3 grams of hydroxpaphthaquinone, 3 C.C. of KaliI- baum's valeraldehyde, and 9 C.C. of alcohol mas boiled for three hours. a reflux condenser being used. A drop of the solution was then diluted with alcohol, and mixed with a weak aqueous solutioil of sodium hydroxide ; the intense crimson colour developed showed that the hydroxynnphthaquinone had undergone change, and, although the cornpouud was not. isolated, the solution undoubtedly contained amylid- enedihydroxynaphthaquinone.* To convert this into its ;I./3-anhydr- ide, the solution was evaporated to dryness i n a dish, on a water orange-coloured substance which lie xas unable to crjst alike.The figures obtained burg, 1883, p. 26). * In the hope of obtaining lapauhol s-ynthetically, I hare studied the behariour of liydroxynaphthaquinone in contact with valerzldehjde under Taried conditions, aiid hare found that if these compounds are heated in acetic acid solution, in the presence of B sufficiently large quantity of hydrochloric acid, an interaction occiirs entirely different from the abore. The resulting compound has been analysed by Mr. C. C. Burger, who has also prepared and analysed its acetjl derivatire. The figures obtained proTe that under these conditions, interaction occurs in the sense of the equation CloH603 + C,HS.COH = C,HI~O, + H,O. needles, melting a t 119-120". and with the alkalis forms violet salts mliich crys- talLise very readily. f l 1 he compound is isomeric with lapacliol ; it crpstallises in brilliant orange-red The acetyl deriratiye fuses at 74"'OF ALDEHYDES WITH b-HYDROXY -#-NAPHTHAQUINOSE.85 bath, to drive off the excess of valeraldehyde, and the residue was dis- solved in 20 C.C. of warm acetic acid, and transferred to a flask which was then immersed in ice-cold water. A mixture of 60 C.C. of Concentrated sulphnric acid and 20 C.C. of acetic acid, also previously cooled, was then added, the flask being kept, during the whole opera- tion, in water cooled by ice ; after a few minutes, it was poured into a large volume of water, and the precipitate which formed was col- lected on a filter and thoroughly washed. When dly, it was twice crystallised from acetic acid and analysed.0,1729 gave 0.4'759 CO, and 0.071 1 H,O. C = 75.06 ; H = 4-56. Cz,H,,05 requires C = 7.5-37 ; H = 4.52 per cent. The anhydride was obtained i n small, orange needles (sometimes also in small, reddish-brown plates), which commence to darken at a temperature near to 200", and fuse and decompose at a somewhat higher temperature. The presence of an orthoquinone group was demonstrated by the behaviour of the compound with orthotolnplenedianiine, the azine being prepared essentially in the manner previously described (p. 81) in the case of the corresponding benzylidene compound. For analysis, it was purified by crystallisation from benzene, in which it is much more soluble than either the corresponding benzylidene or et hylidene compounds. It was obtained in orange-coloured needles which gave precisely the same colour tests with sulphuric acid as the azines already described. 0.1618 gave 0.4705 C02 and 0.0753 H,O. 01653 ,, 8.20c.c. moist nitrogen at 26" and 767 mm. N = 5.55. C32H2rN203 requires C = 79.34 ; H = 4.95 ; N = 5.23 per cent. Cumiuic, Xalicyl ic, aud Ciwiamic Aldehydes. C = 79.30; H = 5-17. Tbese nldehjdes and hydroxpnaphthaquinone interact very rcadily (compare Schoch). By following the methods already given, we have succeeded in converting the resulting compounds, apparentIy quantitatively, into the ap-anhydrides. In all three cases, the anhydrides proved to be bright red, crystalline compounds, which were readily acted on by orthotoluylenediamine, giving characteris tic azines. The compounds were not analjsed, and for that reason are not described mere fully here. I shall defer for a short time the discussion of the constitution o€ the compound and its bearing on that of lapachol; but I may state that I hare succeeded in converting it by a number of interesting changes into compounds containing the full number of carbon atoms preeent in lapachol, and which I have also prepared from layschol itself. The behaviour of other aldehydes under s'milar conditions %-ill also be studied.-% c'. H.

ISSN:0368-1645    

DOI:10.1039/CT8946500076

出版商:RSC    

年代:1894

数据来源: RSC

摘要:

XI1 I.-S,/pt7~sis 0 f Pentci mt7i yl<):i WGZ &ox ylic acid, Hem T I E e t 7i IJ le ii ccci 7.1) ox y lie cc c id ( Hcxci 7i y cl 1-0 be 1 z xoic acid) , By E. HA\wRTH, B.Sc., and IT. H. PERKIS: Jun., Ph.D., F.R.S. DURIX the last few years, the behaviour of benzene derivatives on reduction has been the subject of much investigation, and Bneyer's classical researches in this field (,4nnnZeu, 245, 103; 251, 257 : 269, 145) have shown that, nhen subjected to the action of sodium amalgam. the phthalic acids jield =t variety of interesting derivatit-es, the end product being in all cases the siaturated hexahydro-acids formed by the complete reduction of the benzene ring. Although these acids contain a closed chain of six carbon atoms, they have none of the properties characteristic of benzene deriratives, but beha\-e in almost all respects like saturated open chain acids of the fattj- series.0 ther hexahjdrobenzene derivatives hare also been obtained by the reduction of the corresponding benzene derivatives ; but only in a very few instanses have such compounds been synthetically prc- pared from substances belonging to the fatty series. One case of this kind, which has a special bearing on the present investigation, is the synthesis of 1 : 2-met,hylhexamethylenecarboxglic acid (hexa- hydro-a-toluic acid ; Freer and Perkin, Trans., l8S8, 53, 202), which was carried out in the follon-ing manner. The sodium derirative of ethylic mnlonate was digested in alcoholic solution with methj-lpentarneth~-lene dibromide, when reactioii readily took place in accordance with the equation C C I ~ C ~ ! ilzelcci'c clcitl.2C HNa (C 0 0 C2 H5) + C H,*C HBr*CH2.CH2C H2* C H2Br = The ethylic methylhexamethylenedicai~bosylate was converted by hydrolysis into the correspcoding biljnsic acid, which, when heated a t 2OOo, was decomposed quantitatively into carbon dioxide and methyl- hexametbylenecarbox~lic acid. FEI,*CH,-YH*CH, YH2CHz*$lH-CH, + coz. - C H,.CH?* C (C OOH), CH,*CHz.CH*C 0 0 H Ths action of reducing agents on s-toluic acid has not as yet been investigated ; and, therefore, i t bas not been possible t o compare the reduced benzene derivative with the synthetical acid ; some time since, however, Aschan (ihzna7en, 271, 231) succeeded in convertingSYNTHESIS OF PENTAXETHYLENECARBOSTLIC ACID, ETC.87 benzoic acid into hexahydrobenzoic acid, the method used being the following. In the first place, benzoic acid was reduced by excess of sodium amalgam in alkaline solutiou to A?- tetrahydrobenzoic acid ; and, by treating this acid with hydrobromic acid, it was converted into /3-(?)-bromhexahydrobenzoic acid. HeCOOH H*COOH H/\H H,'\HB~ + HBr = HI, HJ,!. ' H2 H 2 Bromhexahydrobenzoic acid is readily reduced by sodium amalgam, yielding hexahydrobenzoic acid (hexameth ylenecarboxylic acid). This important acid is very similar in its propel ties to the methylhtxa- methylenecarboxylic acid prepared synthetically by the method de- scribed above; and it appeared to us that it would be especislly interesting t o endeavour to obtain hexahFdrobenzoic acid syn- thetically, as it would then be possible t o compare the synthetical acid with the reduced benzene derivative. For this purpose, pentamethylene dibromide, Br[ CII,],Br, was re- quired, a snbstrnce which Gustavson and Demjanoff (J.pi.. Chern., 39, 542) had already prepared from pentamethylenediamine by con- verting i t into the corresponding glycol by the action of silver nitrite, and subsequentJy treating the glycol with hydrobromic acid. NH2-[CH2Jj*KH2 HO* [CH,j,*OH Br*[CH2],*Br Pentametlqlenediamice. Pentamethglen?gl~ col. PentamethTlene di- bromide. Before commencing tbe study of the action of this dibromide on the sodium derivative of eth-j-lic malonate, we wrote to Professor. Gustamon, and in reply he informed us that Herr Demjanoff had already instituted experiments in this direction, but did not intend to proceed any further with them, and subsequently Herr UemjanoE not only agreed to allow us t o continue this research, but he also very kindly gave us a detailed account of the results which he had obtained in his preliminary experiments.The yield of pentamethylene dibromide obtained by Gustavson and Demjanoff ( J . p r . Chenz., 39, 542) was very small, only about 8 per cent. of the theoretical, and as we required considerable quantities of it, we, in the first place, made nnmerous experiments with the object of improving the method of preparation, but without success, so that ultimately me prepared our material by a method practically identical with that devised by Gustavson and Demjanoff. The preparation of the dibromide is exceedingly tedious : altogether 1500 grams of tri- methylene bromide were used, the conversion of wliich into penta- metbylene dibromide necessitated mwe than three months' steady88 HAWORTH AWD PERKIN: SYNTHESIS OF work, the ultimate amount of product obtained being only 90 grams.The subsequent study of she actionof this dibromide on the sodium compound of ethylic malonate yielded very remarkable results, which prove, as we believe, that the substance which was supposed to be compnratirelg pure pentamcthylene dibromide is in reality a mixture of @is substance and tetramethylene dibromide, Br*[CH,],*Br, the lattcr constituting, as it appears, as much as 70-75 per cent. of the whole. It is certainly very difficult to understand how tetramethylene di- bromide can be thus produced from pentamethylenediamine; and i t will hardly be possible t o understand this remarkable method of form- ation, until a very careful examination of the action of silver nitrite on the hydrochloride of the diamine has been instituted, as the decomposition seems t o be very complicated.* The reasons for assuming that the dibromide obtained by Gustav- son and Demjanoff's method is a mixture, will be readily understood from the accompanying short sketch of the results obtained in in- vestigating its action on the sodium derivative of ethylic malonate. In alcoholic solution, action takes place very readily on warming, with separation of sodium bromide and fo~mation of an oily, ethereal salt, which, on distillation, yields, besides regenerated ethylic malon- 'ate, two principal fractions, 240-250' (760 mm.) and 270-275" (59 mm.).The fraction 240-250" contains traces of bromine, and, therefore, did not give good results on analysis ; when hydrolysect by boiling with alcoholic potash, however, i t yields rt beautifully crystalline bibasic acid, which, on analysis, gave numbers agreeing sharply with those required by the formula C7H,,04 ; this result was confirmed by the analysis of the silver salt., which has the composi- tion C7H,Ag204. This acid is, therefore, not hexamethylenedicarboxylic acid, the ebhereal salt of which would have resulted from the action of penta- methylene dibromide on the sodium derivative of ethylic rnalonate ; but it contains CH, less than this acid, and is, therefore, probably pentamethylenedicarboxylic acid, the ethereal salt of which would be produced in the following manner.* While this paper was in the press, i t occurred to me that this formation of tetramethylene dibromide might be explained as follows. When trimethylene di- bromide reacts mith potassium cyanide, it may be assumed that the decomposition, to some extent, proceeds thus, Br-[CH2I3*Br + KCN + H20 = CN.[CH2],.0H + KBr + HBs. The cyanhgdrin, thus produced, would then, on reduction, be converted into hydr- oxytetramethyleneamine, NH~*[CHB],.OH, the hydrochloride of whicl,, in contact with silver nitrite, would jield tetralnethjleneglycol, 0E.[CH2]4.0H, tetra- methylene dibromide being formed from this by the subsequent action of hydrogen bromide.-W. H. P.: Jun.PESTAMETHnENECARBOSYLIC ACID, ETC.89 2CH?u’a(COOCIH5)2 + CH2Br.CH2*CET *CH2Br = CHZ-FH2 + CH,(COOC,H,), + PNaBr. C H ’ < ~ ~ , ~ ( ~ ~ ~ ~ , ~ , ) , That the action actually takes place in this manner is proved by the follom-ing considerations. The bibasic acid of the formula C;H,,Ot does not decolorise potassium permanganate i n dilute alkaline solution ; it is, therefore, a saturated acid, and must contain a closed carbon chain; when heated above its melting point’, it rapidly decom- poses, carbon dioxide being evolved and an oily acid prodoced, which distils constantly a t 214-215O. The analysis of this acid and its silver salt prove that it is a monobasic acid of the formula C6HI0O2. and its formation, on the assumption that the bibasic acid was penta- metliylenedicarboxylic acid, may be represented thus CHZ-YHZ CHZ-YH, = CH,< + co,.CH2< CH?-C (c o o H ) ~ CHZ-CH*COGH Pent ame thylenemono- carboglic acid. The properties of this monobasic acid coincide so exactly with those n-liich could ha\-e been predicted for pentamethylenecarboxylic. :wid, that we had no hesitation i n adopting this riew of its constitu- tior ; curiously enough, however, while this research was i n progress and nearing completion, Wislicenus and Gartner ( AniaaZen, 275, 333). sxcceeded in preparing pentamethylenecarbosylic acid by a series of reactions which leave no doubt as to the constitution of their product By the distillation of anhpdrous calcium adipate, these chemists prepared, in thc first place, the ketone of adipic acid (ketopentn- niethylene) , Q H , C H z > ~ ~ + CaCO,.>Ca = yH,-CH?.C 0 0 CH,*CH,*C 00 CHZ-CH, When this ketone is pouwd on to powdered potassium cyanide, and concentrated hjdrochloric acid is added, the hydroxycyanide first pro- duced is hgdrolped and conrerted into a-hydroxypentamethylene cai*bosylic acid. This hydrosy-acid is readily reduced by heating with hydriodic acid and phosphorus in a sealed tube a t 190-195”, and is thus conrerted into pentamethylenecarbosylic acid, The same acid was a?so obtained by converting ketopentamethjlenc H VOL. LXV.90 HAWORTH AND PERKIN: STNTHESIS OF into pentametbplenic alcohol by reduction ; the alcohol, when treated with hydriodic acid, yields the corresponding iodide, which, when digested with potassium cyanide, is converted into pentamethylenic cyanide ; from this, pentamethyleuecarboxylic acid is obtained by hydrolysis.At o u r request, Professor Wislicenus was kind enough to send 11s a small quantityof his acid in order to enable us t o decide. definitely as to its identity with our acid. We converted his specimen into the acid chloride, from which we prepared the anilide, C5HS*COo~'H*C6H5, vhich crystallises from alcohol in a highly characteristic manner, and melts sbarply a t 159-160'. On repeating the experiment under precisely similar conditions, with our pentamethylenecarborryIic acid, we obtained an snilide which crystallised in the same characteristic manner, melted a t 159-160", and on careful comparison was found io be identical with the anilide OF Wislicenus and Giirtner's acid.There can, therefore, be no doubt that the acid C6HI0O2 obtained by us in the manner described above is in reality pentamethylene- cnrbosylic acid. During the course of this investigation the action of bromine in the presence of phosphorous on pentumetbylenecnrboxylic acid was studied, and in this way some very interesting resnlts were ob- tained. If the product is poured into meth-j-1 alcohol, methylic a-bromopentamet bylenecarboxylate, a colourless oil boiling at 122-125" (60 mrn.), is produced; and this, when treated with aqueous potash, yields Al-pen t ame thenylcarboxylic" acid. CII,<CH'-p + BKOH = CH2-CBr*COOCH, CH+H + KBr + CH,OH + H,O. crrz<c 8,-C.COOK Potassiuni pentamethenjlcarboxylate. This acid melts at 119-121", and in many of its properties sho5v-s a marked resemblance to benzoic acid, from which, however, it is sharply differentiated by its instability towards alkaline permanganate solution, which it instantaucously decolorises.When siibjected to the action of bromine vapour, pentamethenyl- carboxylic acid is conx-erted into a saturated dibromo-addition pro- dnct, C,HSBr,O2 ; t'his is evidently dibroniopentaruethj-lenecarboxylic acid produced according t o the equation CHZ-ZH CH+HBr C H,C C 0 0 H CH2<CH,-CBr*COOH ' + Br2 = CH,< * This name is gircn to the acid in order to indicate its connection with the penta- methenyl derivatiyes described by FFislicenue and Gartner (Aizncrlen: 276, 331).PENTAMETHYLENECARBOXTLIC ACID, ETC. 91 By treating hydroxypentamethylenecarboxylic acid with hydriodic acid and phosphorus a t 150°, Wislicenus and Gartner (loc.cit., p. 337) also obtained, in alniost quantitative yield, an acid C6H802, which melted at 120", and was readily volatile with steam. There is no doubt that this acid is identical with pentamethenylcarboxylic acid, and its formation is readily understood, if it be assumed that the hydriodic acid of the strength employed acted simply its a dehydrat- ing agent, thus, the above mentioned chemists, indeed, discuss this possibility, but consider it very improbable that the reactiou proceeds i n this way vithcut, however, giving any cogent reasons against this assumption ; they are, moreover, unable to suggest any formula, other than the above, whicli corresponds with the properties of the acid.The next step was to determine whether by the action of the mixed bromides on the sodium derivative of ethylic malonate anF hexamethylene deril-ative had been formed. I n order to decide t h i s point, the T-arious mother liquors of the pentamethylenedicarboxylic acid were evaporated t o dryness and the residue w a s distilled. The colourless oily distillate, after treatment with permanyanate to remove unsaturated compounds, was very carefully fractionated, and in this way rather more than 3 grams oE a colourless acid, boiling at 231-233", was obtained. It solidified in a freezing mixture, and on analysis gave numbers agreeing with the formula C7HIZO2, s result which was confirmed by the analysis of the silver salt, CiHllAgO,. That this acid is hesahydrobenzoic acid (hexamethylenecarboxylic by the following facts :- 1.It has the same boiling point as Aschati's hexahydrobenzoic acid. 2. It is a saturated acid, since its solution in dilute sodium carbon- ate does not decolorise potassium permanganate. 3. The mixed dibromides from pentamethylenediamine undoubt- edly contain considerable quantities of pentamethylene dibromide, as is proved by the synthesis of azela'ic acid (see below), for the prodnc- tion of which the presence of this dibromide is necessary : as, there- fore, this dibromide is present, it would be most remarkable if by its action on the sodium derivative of ethylic malonate, some ethylic liexamethylenedicarboxylate were not produced ; and the latter, on hydrolysis and subsequent decomposition by heat,, must yield hexa- hydrobenzoic acid (compare Freer and Perkin, Trans., 1888,53,206).E 292 HAWORTH AND PERKIN: SYNTHESIS OF The formation of hexahydrobenmic acid by 'this reaction is further- niore confirmed by Demjnnoff, who (as he states in the description of his experiments which he kind17 sent us) obtained an ethereal salt boiling at 244-255", which, on hydrolysis, yielded a bibasic acid melting and decomposing a t 145-160" ; on analysis, it gave numbers agreeing with the formula of hexamethylenedicarboxylic acid C,H,o(COOH),* Found, C = 55.6; H = 7.2. Theory, C = 55.8; H = 7. This bibasic acid, on distillation, yielded an oily monobasic acid, t,he calcium salt of which contained Walter of crystallisation, and, altei. drying, gave numbers agreeing with the formula (C,H,,O,),Ca : Found, Ca = 13.6 per cent.Theory, 13.6 per cent. According t o Aschan, the calcium salt of hezab ydrobenzoic acid ha.; the formula (CsH,,O?),Ca + 4H,O. There c m be no doubt that the latter was hexahydrobenzoic acid, and its synthesis from pentamethyl- ene dibiaomidc was, therefore, first accomplished by Dernjanoff. Synihesis of AzeZni'c acid. When the product of the action of the mixed dibromides on the sodium derivative of ethylic malonate is fractionated, the tempera- ture rises rapidly after the fraction 240--250°, which has just been described, has passed over, and if the distillation be continued under reduced pressure, a quantity of a thick oil is obtained, boiling at 270-275" (50 mm.). This oil, which constitutes about 25 pcr cent.of the product, is etbylic heptanetetracarboxylate, formed by the action of 1 mol. of peiitamethylene dibromide on 2 mols. of ethylic sodiomalonate. 2(COOC2H,)2CHNa + Br*[CH,],*Br = On hjdrolysis with alcoholic potash, this ethereal salt yields an oily tetrabasic acid, which, when heated at 200°, is readily decomposed i n t o carbon dioxide and a dark brown, crystalline acicl. The latter crystallises from water in glistening plates, melts a t 107", and is i i i ail respects identical with azela'ic acid, as was clearly proved by ;r direct comparison with a sample of the latter obtained by the oxida- tion of Chinese wax. The apthesis of azela'ic acid has not bcreii accomplished previously, and its formation by the action of heat on heptanetetracarboxylic acid is readily understood with the aid of the following equation. (COO CJ&)2CH* [ CH,] j* CH: (C 00 CZH,), + 2NaEr, (COOH)2CH*[CBz]5.CH(COOH), = C OOH*CH:*[ C HZI5*CH2*COOH + 2 C 0,.No trace of any other acid, siiberic acid for example, was formedPENTABfETHYLEKECARBOSYLIC ACID, ETC. 93 i n this reaction. It is very remarkable that tctramethylene dibromide and pentamethylene dibromide should differ so markedly in their behaviour towards ethylic sodiomalonatc : that is, that penta- methylene dibromide shonld yield, besides ethylic hexamethylenedi- carboxylate, also ethylic heptanetetracarboxylate, but that tetra- methylene dibromide under precisely similar conditions should be quantitatively converted into ethylic pentarnethylenedicarboxylate, without a trace of ethylic hexanetetracarboxylate, heing formed.I n connection with this point, i t is interesting t o note that methpl- pent amc th ylen e dibromide, CH3* C HB r* [ C H2] 3-CH2Br, and met b )-1 tetramethylene dibromide, CH3-CHBrfCH,],*CH2Br, behave in a precisely similar manner when heated mi th ethylic sodiomalonate ; the former (Freer and Perkin, Trans., 1S88, 53, 202, 215) yielding ethylic methylhexamethylenedicarboxylate and considerable quanti- ties of ethylic iso-octanetetracarboxylate,* (C 0 0 C,H,),C H* ICH,] ,.CH (CO OCZHS)?, (C 0 0 C,H,),CH* CH (C H3)* [ C H2] ,*CEi( C 0 0 C2H.j) 2, whereas in tbe case of the latter (Colman and Perkin, Trans., 188S, 53, 185) elhylic methylpentamethylenedicarboxjlate alone is formed. 2C HNa (C 00 C,H,), + CH,C HBr*CH2* CH,.CHBr = CHz-7 H-CH, C Hz-C (C 00 C,H,), + CH2(COOC2H5), + 2NaBr. CH*< As we were i n possession of several grams of ethylic heptanetetra- carboxylate, we studied the action of bromine cjn the disodinm deriva- ti\-e of this ethereal salt, in order to determine whether i t were possible in this way to accomplish a, synthesis of a 7-carbon ring.C H2*C H,* CNa ( C 00 C2H,) CHz<CH,*CH2*CNa(C00C2H5)2 4- Br2 = The product,, on hydrolysis, gave an oily acid, which decomposed a t 200°, carbon dioxide being erolved; the oily residue, after a tixe, deposited crystals of azelaic acid ; unfortunatel? we were not able to isolate any other cr-ystalline substance from the mother liquor of these crystals. Experiments on the action of bromine on the disodium derivatke of ethylic iso-octanetetracarboxylate which were instituted some time since by Freer and one of us (Zoc.d.) pave a similar negative result. As, however, it has been shown (Kipping and Perkin, Trans., 1891, 59, 21G-229 ; Wislicenus a2d Meyer, Annulen, 275, 356-366) # Preriously erroneously called ethylic isoheptanetetracarboxylate.94 HAWORTH AND PERKTN: SYNTHESIS OF that a 7-carbon ring is capa$ble of existence, it is very probable that if these experiments were repeated under different conditions a more satisfactory result would be attained. The results of this research prove conclusively that the dibromide produced from the product of the action of silver nitrite on the hydrochloride of pentametliylenediamine contains, besides penta- methylene dibromide as its chief constituent, teti-amethylene dibroniide.It is, therefore, possible that, the hydrocarbon obtained by Gustarson and DemjaEoff (Bey., 24, 4002) by the action of sodium on these mixed dibromides, and which boiled at 35", was not pure penta- metbylene as these chemists supposed, but a mixture of this hydro- carbon with tetramethylene. Wislicenus end Hentzschel (d?inaEen, 275, 327 j prepared a hydro- carbon, which is probably pentamethylene, by reducing an alcoholic CHZ-YH, with zinc and C H 2- CHI' solution of pentametbylenic iodide, CH,< hydrochloric acid ; i t boiIed a t 50*25--50.75", o r about 15" higher than Gustavson and Demjnnoff's product. The boiling point of pentamethylene m a y be calcdated in various ways, as for example, by subhacting the difference between the boil- i D g points of heptnmethj-lene (9s-101") and of hexamethylene (SO" j from that G f the latter the calculated boiling point being thus about 50", a value agreeing with that found by Wiskenus and Hentzschel.It is, however, remarkable that the nnsaturated hydrocarbon penta- methenylene, CH,< CH2-gH prepared by Wislicenna and Girtner which should boil higher than pentamethylene, was found to boil at 45". CH2-CH' Action of Silcer Nitrite o n Peiztnnzetl~1/7eizedini,riiie Hydrochloride. Gustavson and DemjanoE ( J . pr. Cliem., ii, 39, 542), who first studied t h i s decomposition, obtained as a result a substance boiling a t 162" under a pressure of 30 mm., which they concluded wa3 peilta- methylene gljcol, C5H,,(OH)2 ; by heating this with fuming hydro- bromic acid in sealed tubes at loo", they prepared a dibromide boil- ing at' 204-206", which, on analjsis, gave numbers agreeing with thc formula C,H,,Br,. The yield of the glycol and bromide was, however, but small, 17 grams of the latter being obtained from SO grams of pentamethyl- enediamine, whereas, according to theory, the yield should h a ~ e been 180 grams.As it seemed likely that in the course of our experiments we should require considerable quantities of this dibromide, we made in the first place, numerous experiments with the object of improvingPEXTA3iETRPLENECBRBOXYLXC ACID, ETC. 95 t h e yield, using silver nitrite, sodium nitrite, free nitrous acid, &c., under a great variety of ccnditions, but the results were unsatisfac- tory.Ultimately, we found it best either to follow the method of Gustavson and Demjanoff exactly, or, in any case, to introduce only very slight modifications. The method of procedure finally adopted %--as the followiny :-To a fairly strong solution of pentamethylmediamine hydrochloride rather more silver nitrite was added than the amount re- quired by theory. The nitrite was made into a thin paste with water and added little by little to the solut.ion of the hydrochloride, the mixture being kept well cooled during the addit,ion ; care is necessary in per- forming this operation, as, owing to the evolation of nitrogen, iho mixture froths very much, especially i f the solution be too concen- trated. Next day the mixture was heated on the water bath in a re- flux apparatus for one hour ; the precipitated silver chloride was then filtered off, washed with a little water, and the combined liquors con- centrated by distiilation in a flask connected with a fractionating column so as to prevent the glycol beingcarried over with the steam.I n this operation, a small quantity of oil of rery unpleasant odour passed over with the water, but was not obtained in quantity suffi- cient for further examination. The residual liquid, after. the bulk of the water had passed over, was made into a thick paste with anhydrous potassium carbonate, and the glycol which separated as an oil was removed by extracting six times with ether. On distilling off the ether, a considerable quantity of a brown oil remained, which was alkaline and smelt of ammonia.This was mixed with 10 times its volume of fuming hpdrobromic acid (saturated a t oo), which at once acted on it, producing a hissing sound and dense, white fumes, a considerable amount of heat beitig generated. The mixture was heated on the water bath in a flus3 attached to a reflux condenser for about two hours, then resatnrateri with hpdrogen bromide, sealed up in tubes, and heated in boiling xater for f o u r hours. When cold, the contents of the tubes were poured into water, the heavy oily product extracted with ether, the ethereal solution washed w i t h water and sodium carbonate solution, and dried over calcium chloride. On distilling off the ether, a dark brown liquid remained which had the odour characteristic of the higher dibromides in the fatty series.The bromine determination was made by Carius' method. r3.2398 gare 0.3925 AgBia. C5H,,Br2 requires Br = 69-56 per cent. This result agrees closely with that obtaiued by Gnstavson and Demjnnoff, the mean of their analyses giving Br = 70.1 per cent. spite of the close agreement of the ansl$cal numbers w-ith BY = 69.63.96 HAWORTH LYD PERKIN: STNTHESIS OF those required by the formula Br*[CH,],*Br, the result of our sub- sequent experiments with this bromide proves that it, is not by any means pure pentamethylene dibromide, bnt contains only approxim- ately 25 per cent. of the latter, the remainder being tet>ramethylene dibromide, Br*[CH2I4-Br ; a mixture of this kind contains 72.S9 per cent. of bromine. As it is quite usual €or a dibromide such as this, owing to slight de- composition and consequent loss of hydrogen bromide, to gire num- bers from 1 to 2 per cent. below the theoretical, it is not surprising that the numbers found in the present case are also inaccurate.Action of the JIixeJ Tetrainethylel2e a i d Peiltam ethyle,ie Dibromides O I L tlrr Sodium Derlcatire of E:fhylic Xzloiinte. 8 grams of sodium were dissolved in 100 grams of absolute alcohol, the solution well cooled, and a mixture of 40 grams of the bromide and 56 grams of ethylic malonate added. At ordinary temperatures. no change appears to take place except on long standing, but, on gently warming, the action soon begins, sodium bromide separates, and the liquid boils vigorously for some time; as so011 as the reaction becomes sluggish, the mixture is heated on the x a t e r bath in a reflux- apparatus for about two hours, allowed to cool, poured into twice its volume of water, and then extracted with ether four times.The ethereal solution is well washed with water t o free i t from alcohol, dried over calcium chloride, and the ether distilled oE. Two experi- msnts mere made in this way, the total amount of bromide used being 90 grams. The products from both these experiments were mixed and frac- tionated under reduced pressure (50 mm.), and in this way separated into three fractions boiling (1) between 150" and 145" ; (2) between 15Y and 180" ; and ( 3 ) between 180" and 2i5". A small portion of fraction 2, which boiled a t 162-165" under a pressure of 30 mm., was collected for analysis, but i t did not gire satisfactory numbers, om-ing to the fact that it contained traces of bromine.The two fractions 1~-30-180" were then mixed, and distilled under ordinary pressure, when they were separated into tn-o parts, the oue boiling below 210" and consisting almost entirely of ethylic malonate, the other a thick oil boiling a t 210-250". The further purification of fraction 3 boiling at 180-%5"(;0 mm.) \rill be described later (p. 104). CHZ-YH, 1 : 1 -Pe~~tameth?llenedicar~ox~l~c acid, CH,< CH,-C (CO OH),' The fractions of the products of t h e action of the mixed bromides on ethylic sodiomalonate boiling at 200-210" and 210--250" cannotPE?JTAMEl'HTLENECARI30XTLIC ACID, ETC. 97 be readily purified by fractional distillation, Gwing to the fact that the oil contains bromine which cannct readilF be removed, and the presence of which renders the results of analysis valueless ; for t h i s reason, the oils were directly submitted to hydrolysis and the resuit- ing acids further examined.Fraction 200-2 lo", consisting principally of ethjlic malonate, and weighing 37 grams, was mixed with a strong solution of 40 grams of pure caustic potash in methylie alcohol, and heated in a reflnx apparatus for two hours. The alkaline solution was then diluted with water, evaporated on a water bath till free from alcohol, and the residue cooled, acidified with dilute sulphuric acid, and extracted five times with pure ether. The ethereal solution, after drying over calcium chloride and evaporating, deposited a small quantity of crude pentamrthj1enedicarbox)-lic acid, the malonic acid remaining almost entirely in the aqueous solution.This was subsequently worked up with the principal quantity obtained from the fraction 210-250". During the l~jdrol~-sis of the fraction 210--250", weighing 55 grams, with 40 grams of potash dissolved in methS-lic alcohol, a considerable ilnantitF of a cyxtalline potassium salt separated. The whole dis- solved readily in water, xiid on evaporating, acidifying, and extracting with ether, as described above, 4.L grams of a brown, oily acid was ob- tained, which, on standing over sulphuric acid in a vacuum, deposited crystals, and gradually became semi-solid. This crude product was spread on biscuit ware, and when the brownish, oily mother liquor Ilad been entirely absorbed, the colourless, crystalline mass which remained was recrystallised several times from water.I t then formed coionrless prisms, which, on analysis, gave results agrecing with the formula of pentamethylenedicarbox~-l~c acid, C5Hs( COOH),. 0.1570 gave 0.3060 C 0 2 and 0.0910 H,O. C = 53-16 ; H = 6 44. 0.1281 ,, 0.2493 ,, ,, 0.0752 ,, C = 53-08; H = 6-33. C,Hs(COOH), requires C = 53.16; H = 6.33 per cent. 1 : 1-Pentamethjlenedicarboxylic acid is readily soluble in hot, but comparatively sparingly in cold, water. It crjstallises from water in colourless prisms, which, --lien heated in a capillary tube, melt a t about 184-185", undergoing decomposition into carbon dioxide and pentamethylenemonocarboxylic acid ; this decomposition also takes place to some extent when the aqueous solution of the acid is evaporated, as the solution acquires a strong odour of the monocai-b- oxplic acid.It is readily soluble i n alcohol and ether, sparingly in benzene, and almost insoluble in light petroleum. The solution of this acid in sodium carbonate does not decolorise potassium perman- canate, even on long standing, showing that the acid is saturated. Xilcer Salt of 1 : l-Pe~ztumethyle~zedicurbox.ylic acid.--This was ob-98 HAWORTH AND PERKIN : SYNTHESIS OF tained as a white, amorphous precipitate, on adding silver nitrate to a neutral solution of the ammonium salt of the acid; it is very sparizgly soluble in water. After being well washed, it was dried over sulphuric acid in a vacuum and analysed. C = 22.19 ; 0.2895 gave 0.2418 COz, 0.0551 H,O and 0.1668 Sg.0.3157 gave 0,1816 Ag. H = 2.11 ; Ag = 57.66. Ag = 57.52. C,Hs(COOAg), requires C = 22-58 ; H = 2-15 ; Ag = 58.06 per cent. CHZ-YH:, Pentam eth y lenenonocarboz y lic acid, CH,< C H,-CH*C 00 H' When pure pentamethylenedicarboqlic acid is heated in a dis- tilling flask, carbon dioxide is evolved, and an oil consisting of pure pentameth ylenemonocarboxylic acid distils over constantly at As the quantity of pure dicarboxylic acid at our disposal was small, we extracted, with ether, the biscuit ware containing the crude oily dicarboxylic acid (p. 97), evaporated the ethereal solution, and decomposed the brownish residue, by heating it at 200" until no more carbon dioxide was evolved. On distilling the residual oily acid, the whole (20 grams) passed over between 210" and 230" as a very unpleasant-smelling oily liquid which did not solidify when cooled in a freezing mixture.This acid is a mixtiire of pentamethylene- carboxylic and hexamethylenecarboxylic acids, the former being pre- sent in by f a r the larger quantity. The acids are, however, con- taminated with some impurity which causes the alkaline solution to decolorise permangannte solution in the cold. In order t o remove this impurity, the mixed acids were dissolved in dilute sodium carbonate solution, cooled below O", and perinangan- ate solution (1 per cent.) added until the pink coloiir was permanent. The excess of permanganate was then removed by adding a few drops of alcohol, the product filtered, and the filtrate evaporated to a small bulk.The acids were then liberated by adding dilute sulphnric acid, extracted with pure ether, and the ethereal solution, dried over calcium chloride, was evaporated. The residual oil (17 grams), sub- mitted to very careful and repeated fractional distillation, gave about 9 grams of oil boiling constantly at 214-215" (the higher fraction is mentioned later), which, on analysis, gave results agreeing with the formula of pentamethylenemonocarboxylic acid. 214-215". 0.1998 gave 0.4612 GO, and 0.1585 H,O. The silver salt, CsH902Ag, obtained by adding silver nitrate t o a iieatral solution of the ammonium salt, as a voluminous, white, amor- C: = 62.95 ; H = 8.80. C,H:,COOH requires C = 63.16 ; H = 8.77 per cent.PENTXI\IETHYLENECARBOSTLIC ACIDt ETC.99 18 * 5 O 18 * 5 18'5 18-5 18 -5 Average 18.5 plious precipitate, was well washed with water, dyied over sulphwic acid in a vacuum, and malysed. C = 32.41 ; 0.1496 gave 0.1778 COs, 0.0583 H20, and 0.0730 dg. 0.1872 gave 0-0908 Ag. H = 4-32 ; Ag = 48.79. Ag = 48.5. C5H9COOAg requires C = 32-58; H = 4.0'7 ; A g = 48.8G per cent. The relative density, magnetic rotation, and refractive power deter- minations were carried out by Dr. W. H. Perkin, sen., who obtained the following results. Pentamethyl e fiemonocarha y Eic acid. The density determinations gave- a 40140 1.0540. t z 15"/15" 1.0452. cZ lO"/lO" 1.0489. d 20"/20" 1.0916. d 2.5'/25" 1.0385. The refractive power determinations gave- t. PA. re PD. r" F' PG- 17.7" 1.44759 1*45040 1.45280 1.45858 1.46314 0 '9633 0 -9T08 0 '9738 0 -9669 0 '9753 0 -9700 ----- P - 1 T P .48.966 49.273 49.536 The magnetic rotatlons gave- t. I sp. rotation. 50,168 50.668 Mol. rotation. 5'851 5 -896 5 -914 5 *873 ti ,924 5 *891 As in the case of all the saturated closed carbon chain carboxylic acids which have so far been examined, the magnetic rotation is very low. Senta?nethzJleiaecarbox~l~c Chloride, C51T9*COC1. This was prepared by heating the pure monocarboxylic acid (2 grams) with twice its volume of phosphorus trichloride for about 10 rninutcs, and after allowing the mixture to stand for about an hour, decanting the liquid from the layer of phosphorous acid, and submitting it to fractional distillation. As soon as the phosphorus trichloride had passed over, the thermometer rose rapidly to 160°,100 HAWORTH ASD PERKIN: SYNTHESIS OF between which temperature and 162O the whole distilled.This fraction was analysed by Carius' method, 0.1571 gave 0.1652 AgCl. C1 = 26.09. C6H90C1 requires C1 = 26.79 per cent. This chloride has a very similar smell, and is generally I-ery similar in properties to the corresponding chloride of tetramethylenecnrh- oxylic acid $l Hz--YHz which boils a t 139". CH,--C K*CO C1' A?&& of Pe,ifan2etl~7_/7e?teL.ar-2,omy7ic acid, C5H,.CO*NH*C,H,.-In order to prepare this, the chloride of pentameth~lenecarbox~-lic. acid was added drop by drop t o a large excess of pure aniline, the whole being stirred with a glass rod during the operation. As soon as the rigorous action had moderated, the mixture WLS heatetl on a water bath for a fen- minutes, poured into water, and dilute hydrochloric acid added to dissolve the excess of aniline.The re- sulting crjstalline precipitatc W:LS well washed w i t h water, dried on a porous plate, recrystallised several times from alcohol, and anal?-sed . 0.1975 gave 13.1 C.C. moist nitrogen at 1'7" and '762 mm. X = '7.i1. C,H,-CO*NH*C,H, requires N = '7.40 per cent. The anilide of pentamethylenecarboxylic acid crystallises from alcohol in magnificent glistening prisnis, resembling s ugnr in appear- ance. It melts a t 159-160", and when strongly heated distils apparently without decomposing. It is readily soluble i n alcohol, benzene, chloroform, and acetic acid, less so in ether and light petr- oleum. Professor Wislicenus was kind enough to send us a small quantity of pure pentamethylenecarboxylic acid prepared €rom the ketme of adipic acid (p.89). I n order to determine whether this was iden- tical with our acid, we converted it into the acid chloride by treat- ment with phosphorus trichloride, and prepared from the fractioneci chloride the anilide under exactly the same conditions as those given above. The anilide, after two crystallisations from alcohol, melted at 159-160", and give the following results on analpis. 0.2200 gave 13.9 C.C. moist nitrogen at 17" and 766 mm. N = 7.39. C,H,.CO*NH*C,H, requires N = 7.40 per cent. A very careful comparison of this product with the aniiide obtaineci I-om our acid proved that the snbstaiices mere iden tical.PENTAMETHFLENECARBOXYLIC ACID, ETC.1101 Met l y Zic a- ,XonoEro,mope la tame t IL y 1 e ti ecar b ox y 1 J t e, The action of bromine on pentnmethylenecarboxylic acid was studied in the first instance, with the view of showing that, when treated in this way, the acid behaved as a saturated acid, forming a monobromo-substitution and not a dibrom-additive product. 5 grams of pentamethylenecarbox~lic acid (b. p. 214-217') was mixed with 0.3 gram of dry amorphous phosphorns in a flask connected with a retlnx condenser, and 15 grams of dry bromine added in small portions a t a time. Ths decomposition took place at once, quantities of hydrogen bromide heing evolved. As soon as the violence of t h e action had moderated, the whole was heated on a water bath for five hours, the condenser then removed, and the heating continued untii' the excess of bromine had been driven off.The dark-coloured product was converted into the met,hylic salt and not into the acid, as it was thought that the former would be more readily purified. For this purpose, the oil mas added to excess OE methylic alcohol, and after standing f o r an hour, the alcoholic solution was ponred into water. The oily methylic salt was now ex-. traeted with ethcr, and the ethereal solution, aiter being well washed with water aud dilute sodium carbonate solution, waa dried over anhydrous potassium carbonate. On distilling off the etber, a dark brown oil was left, which, on being purified by fractionation under reduced pressure, yielded a colourless oil ; this, on analysis, gave t h e following results.0.3203 gave AgBr = 0.2960. Br = 39.3. 0.2326 ,, AgBr = 0-2165. Br = 38.8. C6HiBr*C90CH requires Rr = 38.8 per cent,. ,Methy& a- bi-onzcpentawetlr y1enecarb:lxyEnte is a colourless oil which boils at 122-125" (60 mm.), and is specifically heavier than water, Tn i'tc, odour and general properties it closely resembles the come- sponding tetramethylene derivative (Trans., 1892, 61, 43). CHZ-GH CHZ-C*CO OH' A'- Pc3ntametlcenylcarbozylic acid, CH,< Tllis in tercsting unsaturated acid is formed when methylic mono- brornopentamethylenecarboxylate is acted on by aqueous potash, as explained in the introduction (p. 90). I n order to prepare it, methylic monobromopentamethylenecarboxylate was mixed with strong potash solution (sp. gr. 1.2) and shaken from time to time until i t had alniost entirely dissolved, an operation which took about102 HAWOKT-J AXD PERKIX: SYNTHESIS OF two days.The solution was t.hen heated to boiling for half an hom, cooled well, acidified, and extracted five times with pure ether. The ethereal solution was dried with calcium chloride, and evaporated, when a light brown oil was left, which, after a time, solidified almost entirely. The crystals were freed from oily impurity by spreading the mass on a porous plate, and recrystallising three times from water t o which a little purified animal charcoal was added. I n this way beautifnl, colourless crystals mere obtained. 0.1352 gave 0.3144 C 0 2 and 0.0870 H20. C,H,O, requires C = 64.2s ; H = 7.14 per cent. The annl-j-sis and general properties of this acid, and especially its bchaviour when treated with bromine (see below), leave scarcely any doubt that it is A'-pent~methenylcarboxyli~ acid.When heated in a capillary tube, the acid softens a t about 215", and melts not quite sharply at, 119-121". It is readily soluble in hot n-ater, alcohol, ether, and light petroleum, but only sparingly iu cold water. It dissolves easily in alkalis and alkali carbonates, the cold solution in dilute sodium carbonate decolorising permsnganate instantaneouslJ-, showing that the acid is unsaturated. Pentamethen-j-lcarboxylic acid crjstallises from its solution in hot water or hot light petroleum in glistenizlg plates, which yery closely resemble benzoic acid in appearance. It is also remarkable that not only are the melting points of the two acids identical, but also that pentamethenjlcarboxylic acid sublimes with great rcadi- ness even a t 100" in T-ery much the same way as benzoic acid.Pentamethenylcarboxylic acid is, without doubt, identical with the acid C,H,O,, of which Wislicenus and Gartner (Aiznalen, 275, 337) obtained an almost quantitative yield by heating hydroxy- pentamethjlenecarboxylic acid with feebly fuming hj-driodic acid and phosphorus a t 150" (see p. 91). C = 64-22 ; H = i.25. CH+HBr C Hz-C BY C 00 H. Dibi-onzopeiztnmetTiylcnecarbox7J7ic acid, CH,< In order t o obtain further evidence in support of the view of the constitution of pentamethenylcarboxrlic acid adopted in the pre- ceding section, the action of bromine on this acid was carefully inves- tigated.0.06 gram of the pure acid was placed on a watch glass under a bell jar, and exposed to the action of bromine vapour for 10 hours ; the excess of bromine was then remored by allowing the dark brown liquid product to remain over potash in a vacuuni for 24 hours. I n this way an almost colourless, crystalline mass was obtained, which, after recrystallisation from light petroleum, gave, on analysis, numbers agreeing with the formula of dibromopentametbylene- carboxylic acid.PENTAMETHTLENECXRBOIYLIC ACID, ETC. 103 0.0984 gave 0.1358 AgBr. CcH8Br,02 requires Br = 58.82 per cent. Dibromopentamethylenecarboxylic acid is readil1 soluble in ether, alcohol, chioroform, benzene, and hot light petroleum, sparingly so in the latter solvent in the cold, and almost insoluble in water.It crjstal- lises from light petroleum in colourless leaflets, which, when heated in a capillary tube, soften a t 1 2 7 O , and melt a t about 134". A freshly prepared solution of the acid in sodinm carbonate does not decolorise permanganate, but it does so after standing for some time owing to slight decomposition. When boiled with water, the acid first melts, and then dissolves completely, the solution now containing much hydrobromic acid. Br = 58-81, When crude peiit~rnetb~lenedicnrboxylic acid is distilled, and the oily distillate repeatedly fractioned, the principal product obtained is pentarneth~lenecarbo~~li~ acid, boiling a t 214-215". But there is also a considerable quantity of a higher fraction, boiling between 215" and 23.5" ; this: n - h e n submitted t o repeated and careful fractionation, ya-re ahout 3 grams of a colourlcss oil, which distilled between Z32' and 234".On analpis, it gave the following numbers. C = 6 . 3 7 ; H = 9-39. 0.1320 gave 0.3164 COz and 0.1116 H20. C,H,,.COOH requires C: = 65.62 ; H = 9-38 per cent. The si7rer salt of this acid was obtained as a white, amorphous pre- cipitate on adding silver nitrate to the neutral solution of the am- monium salt; after washing i t well with water, and drying 01-er sulphuric acid in a vacuum, it was analjsed. C = 35.45 ; C6H1,*COOAg requires C = 35-74 ; H = 4.69 ; Ag = 45.95 per cent. These results agree closely with those required for hexahydro- benzoic acid, and there is every reason to believe that the t w o are identical.Aschan (dnfzalen, 271, 260), who prepared hesahydro- benzoic acid from benzoic acid, gives t h e boiling point of the former as 2:32-233", and the melting point at 23", and states that the acid does not decolorise potassium perrnanganate in dilute sodium carbon- ate solution. Our acid distilled a t 232-234", and was also indifferent t o permanganate. When cooled in a freezing mixture, it solidified, lmt gradually liquefied again at ordinary temperatures. This diEer- ence in melting point is obviously due t o the presence of some slight 0.2010 gave 0.2613 C02, O.OE68 H20, and 0.09% Ag. H = 4.79 ; Ag = 46.0'7.104 HAWORTH AND PERKIN: STNTaESIS OF impurity ; but, owing to the small amount of material at our disposal, we were not able to remove it. E thy& Hep tane- wz w ,- te tr aca r b ox y 1 ate, (COOC,~~)2CH*[CH~]j~CH(COOi=,Hj)2. I n fractioning the product of the action of the crude dibromidesor: ethylic sodiomalonate, a thick oil ivas obtained boiling a t 180-285" (50 mm.), and from this, on reciistilliiig, an almost colourless fraction boiling at 270--975" (50 mm.) mas isolated.This, on analysis, gave the following results. C = 56.24 ; H = Salci. 0.1758 gave 0.3754 CO, and 0.1391 H20. C,,H,,O, requires C = 38.76 ; H = 8-25 pel- cent. This substance is etbylic heptane-w,w2-tetracarboxylate, formed bj- the action of pentamethylene clibromide on 2 mols. of ethylic sodiu- malonate, as explained in the introduction (p. 92). It is an almost colourless oil which appears to undergo very slight decomposition when distilled under the ordinary pressure.When added to an ethereal solution of sociiurn ethoside, a yellow, flocculent sodium derirat ive is precipitated ; this probably has the constitution (COOC,H,)?CNs.[CH,],.CNa(COOC,H,)?. HyJroljxis of Ethylic Hepta7ietet,,ncn1.tosylate. acid. COOH. [ C'H,!;.COOH. This ethereal salt was readily hgdrolysed by boiling with excess OF alcoholic potash, the action being complete after heating for half ail hour on a water bath. I n order to isolate the product, water was added, the solution evaporated on :L water bath until free from alcohol, acidified, and extracted several times with pure ether. The ethereal solution after drying over calcium chloride and evaporating deposited a thick, pale yellow oil, probably heptanetetracarbosylic acid; it did not crFstallise, and therefore was not analpsed.When heated at 200", this thick, oily acid readily decomposed, with erolution of carbon dioxide ; the residual oil, which became nearly solid on standing, was readily puri6ed by dissolving i t in boiling water aod decolorisiug by mems of animal charcoal. Beautiful, colourless crystals separated from the hot concentrated solution as i t cooled ; after recrystallisation, it melted a t 105-107" Its identity -&h azelaic acid was established by a direct comparison with a sample of the latter, prepared by the orciclation of Chinese wax. The analysis gave the following results. Syntltesis of dzeluic 0.1102 gave 0*2:301 CO, and 0.842 H,O. C = 57.03 ; H = S.49. C9HL60, requires C = 57-44 ; H = 8.51 per cent.PENTAMETHYLENECARBOXTLIC ACIT), ETC. 105 The siEver salt, prepared from the neutral solution of the ammonium salt by precipitating with silver Eitrate, is a white, amorphous powder, and gave the following results on analysis. C = C = I. 0.1298 gave 0.1240 COz, 0.04534 H30, and 00686 Ag. 11. 0-2296 gave 0.2232 COz, 0.0740 H20, and 0.1230 Ag. 26.06; H = 3.75; Ag = 53.60. 26-53; H = 3.58; Ag = 53-57. C9H,,0,Ag2 requires C = 26.86 ; H = 3-48; 82; = 5373 per cent. Action of Bromine on the Disodiunz Derirative of EthLilic Heptanetetra- carboz y lat e. The action of bromine on the disodium derivative of ethylic iso- octane te tracarbox y late, (COOC2H5)2CNa~CH(CH3)~[CR2],~CNa(COOC,H,)z, was studied by Freer and Perkin (Trans., 1888, 53, 220), with the object of preparing a heptamethylene derivative, but no such sub- stance could be isolated ; a similar want of success has attended our experiments on the action of bromine on the disodium derivative of ethylic heptanetetracarboxylate. I n carrying out this experiment, ethylic heptanetetracarboxylate (1 mol.) was mixed with an ethereal solution of sodium ethylate (2 mols.), and the calculated quantity of bromine (1 mol.) was gradually added, the whole being well agitated and cooled during the addition. The whole of the bromine disappeared, and the action was then evidently not a t an end, as on adding iodine a considerable quantity of this was also absorbed ; in fact, the action evidently pro- ceeds in a very different way from that which takes place when the disodium derivatives of ethylic butane- and pentane-tetracarboxylate are acted on by bromine. On hydrolysis, the prodnct gare an oily acid, and this, when heated at 200°, yielded an oil from which a small quantity of a crystalline substance was obtained melting a t 106", and having all the properties of azela'ic acid. The oily products mere not further investigated. Chemical Laboratories of the Owens College, Munchester. TOL. JrXV. I

ISSN:0368-1645    

DOI:10.1039/CT8946500086

出版商:RSC    

年代:1894

数据来源: RSC

摘要:

106 X1V.-The Oxides of the Elements uizcl the Law. By R. M. DEELEP, F.G.S. LAST session I had the honour to communicate to the Chemical Society a paper describing “ A New Diagram and Periodic Table of the Elements’’ (Trans., 1893, 63, 852). Since this was written, my attention has been called to R very interesting paper, by Professor K. W. Zenger, of Prague, published in the Sitzu?,gsberichte konig. Bohm. Ges. Wissenschaft. (1881, p. 408). I n this communication the author points out that the products obtained by multiplying the specific heats by the densities of the elements have values which are very similar for many elements having analogous pro- perties. To show this, he arranges the elements in I1 groups, the first group containing bromine and iodine, and the eleventh manganese, iron, cobalt, and nickel.Of course the result is only partially satis- factory, for we now know that elements belonging to the same classes have volume heats which, on the whole, decrease with increasing atomic weight, end that even series elements must be separated from odd series elements, although their volnme heats may be nearly iden- tical. I now purpose laying before the Society the outcome of some further work undertaken with a view to ascertain t o what extent a Consideration of the oxides supports or modifies previous conclusions ; for although in my previous paper some reliance was placed on the oxide-forming power of the elements, the diagrams were almost wholly based upon other periodic phenomena. Of course it is im- probable that any one periodic table can express the whole truth, for elements closely resembling each other in some respects may differ widely i n others.Still there does seem to be a fundamental periodic basis, in accordance with which the elements may be classified, b u t several other derived forms may be drawn t o illustrate the periodicity of the several physical or chemical properties. ~ o l u m e heats and volume atoms) mere plotted as ordinates upon a diagram, the abscissce of which were atomic weights. So regular were the curves and lines obtained by joining the summits of the ordinates, that it was possible to make use of the diagram for the purpose of separating the elements into classes and series, and, therefore, for constructing a periodic table based upon direct numvicrtl values. Lothar Meyer (Xodern Theories O J Chemistry, English Translation, tj 63) very clearly pointed out how such a diagram may be inter- I n my previous paper, certain valuesOXIDES OF THE ELEMENTS AND THE PERIODIC LAW.107 preted. He found that, as a general rule, elements having similar properties occupied similar positions on the several curves, and also that each wave of the diagram constituted a complete period. So useful has this method of interpreting the atomic diagram proved to be, that we may, I think, regard with some confidence the periodic variations of the ordinates with increasing atomic weight as valuable and trustworthy aids in investigations relating to the Periodic Law. Unfortunately our knowledge of the relative densities and specific heats of the elements, and even oE their compounds, is by no means complete.In the case of many elements, no determinations have been made, whilst other elements are known to exist in more than one physical condition, and to have two or more specific heats or densities. The fact that there is sometimes more than one value for the same element does not, however, seriously interfere with the symmetry of the volume heat diagram. The diagram Lothar Nejer had pi-eviously drawn had atomic volumes plotted upon it. He also interpreted the curves obtained in accordance with the periodic table drawn up by Mendelkeff. Being, however, led by the volume heat diagram to somewhat modify this table, especially as regards the grouping of the elements of low atomic weight, I have attempted to further elucidate the matter by plotting upon diagrams some other values Calculated from the relative densities of the oxides and the atomic weights of the elements. Although Brauner and Watts (Ber., 14, 48) have already proved that the niolecular volumes of the oxides do show periodic varia: tions, there do not appear to be any methods other than such as may be regarded as purely empirical, of diagrammatically illustrating the periodicity of the oxide-forming power of the elements and the periodic changes which the relative densities of the oxides undergo.One such method I have followed in constructing Diagrams 1: 2, and 3, (p. 115) which show some interesting results. -- Relative densities of oxides Atomic weights of elements- On these diagrams the ordinates = If t,he combination of the elements with oxygen produced com- pounds having densities which were the mean of their constituent atoms, ordinates calculated in this manner, and plotted upon a, diagram, mould give results similar to those of a volume atoms diagram.But the densities of the oxides do not follow such a law, and it is on this account, that the ordinates calculated in the above manner are interesting. Still, there is frequently a close agreement between the periodic variations in the densities of the elements and of their most stable cxides. In my previous paper, (Zoc. cit., p. 864) the elements were arranged in duplicate in one table, the odd and even series, 1, 2, 3, &c., giving108 DEELET: THE OXIDES OF THE ELEMEKTS what may be called the fundamental basis of the Periodic Law, and the duplicates l a , 2a, 3u, &c., actual or possible departures from it.The fundamental basis is shown in Table I (p. lll), and a special form in Table II (p. 112). In the third series, there are ten substances, all of which are regarded as elements. If there are really 10 elements in this series, there should also be 10 in the first, fifth, &c. ; but rather than adopt this view, I would suggest that either one of the substances in Series 3 is not an element, or if it be an element, its atgmic weight is not correctly given. Although Xi and Co have been placed in the same group, it is qnite possible that they are both true elements. Omitting the " typical element " fluorine, Mendelheff, in his table, makes man- ganese the only member of the left-hand column of his seventh group, spaces being left for the undiscovered elements required to complete the table as far as t'his group is concerned. It would, therefore, per- haps, be better to regard either manganese or iron as the redundant element.Diagrams 1, 2, and 3, although they do not furnish any evidence bearing on this particular point, show much the same general features as does the volume heat diagram, and in some respects confirm, or lather give additional support to, the fundamental periodic basis shown in Table I. The vertical scale of Diagram 1 Bas been made smaller than that of the other two diagrams, so as to admit of the ordinate of hydrogen being plotted upon it. Although I have drawn in curves and lines to render clear the general nature of the variation in the magnitude of the ordinates with increasing atomic weight, it must not be supposed that the variations from these cnrves or lines are necessarily due to illcorrect determinations of the densities, $c.Indeed there seems to be evidence that the lines and curves should have been drawn in duplicate, one set cut'ting the ordinates of the perissads, and the other those of the artiads. For instance, the perissads boron, nitrogen, aluminium, and phosphorus all stand above the curves, whilst the artiads carbon, silicon, and sulphur fall below. In Table 111 (p. 113) are given, A, the elements; B, the atomic weights of the elements ; C, the densities of the elements ; D, the oxides; E, the densities of the oxides; F, the volume atoms of the elements ; and G, the laat column, E t B.Diagram 1 has plotted upon it the values in column G correspond- ing to those elements haTing atomic meights less than that of potas- sium. The ordinate plotted for hydrogen is calculated from the density of ice, whilst for lithium it is calculated from the density of Li20, and so on. As on the volume heat diagram, the ordinate for lithium is shorter than that of hydrogen. The sniall numerals on theR.H . v E €LEY. -Tourn. C'hem. SOC. 226.13'94 DIAGRAM I II ATOMIC WEICHT- 4 0 5 0 6 0 7 0 8 0 - K C A s c TI V MN FE co NI c u CR Z N C A C E A s S E BR I SPECIFIC GRAVITY OF OXIDE ATOMIC WEIGHT OF ELEMENT .=(.n : 0 d 1 100 i10 120 130 % _ , Re Y SR Z R N B Mo R u Rti P o AC C O I N SN SS T K I SPECIFIC GRAVITY OF OXIDE ATOMIC WElCHT OF ELEMENT e = IAND THE PERIODIC LAW.109 diagrams give the valencies of the oxides plotted. As a general rule, these are the oxides of which the relative densities are well known. Many non-con secu tive oxides, the relative densities of which are known, have been omitted. In Table I (p. Ill), showing the probable fundamental periodic basis, a space occurs for an element having an atomic weight greater than that of hydrogen, and less than that of lithium. Also, owing to the low density of amorphous carbon, and the anomalous properties of the “ typical elements,” it mas suggested in my previous paper that the volume heats of the elements of the first series, for some reason or other, did not occupy positions on a line such as that of Series 3.Sodium and magnesium were also regarded as having departed from their positions on the curve of Series 2. These views receive additional support from the diagram of the oxides, for, omit- ting hydrogen, the ordinates plotted on Diagram 1 fall on two short lines and two curves, the lines and curves following each other alter- nately. On the first line, me have Li,O and BeO, and on the second line, Na,O and MgO. On the first curve fa11 Be20,, COO, and N,O,; whilst on the second curve there are A1303, SiOz, P205, and SO,. The series from lithium to fluorine is, therefore, exactly similar to that from sodium to chlorine, and they should be placed in the same order on the derived Table 11, although the fundamental order is as shown on Table I.However, on the derived Table 11, which has been reproduced from my previous paper, the typical elements have not been tabulated exxtly in this manner, the properties of the pure elements seeming to demand a rather different arrangement. To agree with the diagram of oxides, the following arrangement (p. 110) would be better. I may here call attention to the fact that the space nieasnred on the atomic scale between hydrogen and potassium is shorter than that between potassium and rubidium, and also than that between rubidium and caesinm. No doubt this is in some way connected with the anomalous group properties of the “ typical elements.” Iudeed it might be urged that on this account the elements of the first and second series should not be classed at all with those of the third, fourth, fifth, and sixth.At present our periodic arrangement should not be regarded as more than tentative; hut the accumulation of more data will doubtless eventually enable the true significance of u-hat has already been called the Periodic Lam to be realisecl. Turning now to Diagram 2, on which are plotted the ordinates for the oxides of the elements between chlorine and rubidium, we have, on the first line, K20, CaO, Sc203, and Ti02. But here a change takes place in the value of the ordinates, for V,05 and CrO, have considerably shorter ordinates thaE those of the four preceding oxides. Here Li and Be are in Series 1’. 1 2110 DEELEY: THE O D E S OF THE Series. L, ELEMENTSAND THE PERIODIC LAW. 111 Series of elements.7 -- 7- 3 .a. 4 4 k? 3 ? 4 CL.112 DEELES: THE OXIDES OF THE ELEMENTS 1 e % - G Fi -~ 3 n Y d' Series of elements. 1 L- r- c a. 4 tt I u c ' 3 c c c 1 z i5 .n . + k a. H n 3 r-4 k t- I_ n n 1 I I HAND THE PERIODIC LAW . 113 TABLE IIr . A . -. Elements . Hjdrogen ..... Lithium ...... Boron ....... Carbon ....... Sodium ....... Magnesium .... Aluminium .... Silicon ........ Phosphorus .... Beryllium ..... Nitrogen ...... .......... Sulphur ...... Potassium ..... Calcium ....... Scandium ..... Titanium ...... Vanadicm .... Chromium .... Manganese .... Iron .......... Cobalt ........ Nickel ....... . . . . . . . . .......... .......... Copper ........ .......... Zin :: .......... Germanium ... Arsenic ....... Selenium ...... Strontium .....Yttrium ...... Zirconium ..... Niobium ...... Molybdenum . . Ruthenium .... Silver ......... Cadmium ..... Indium ....... Tin .......... Antimou? ...... Tellurium . . . . / lodine ......... ............ . . . . . . . . . ........... . . . . . . . . B . .. Atomic weights . .- 1 -0 7 *02 9 -1 11 -0 12 -0 14 -03 2.3 04 24 -3 27 *01 29 3 3 31"03 32 -06 39 -14 40 -08 44 *1 48 -13 51"4 52 . 3 55 *0 56 *02 58 *i 58):6 63':1 65 -3 7, 7i13 79):0 75 *09 87 -5 89 -0 90 -0 94-0 95 -7 101 -65 107 -93 112'11 113 -7 119 -1 1.20 *3 12: -0 126 *83 ............. C . I D . Densities . 1 Oxides . I 0 *62 0 *59 1 -85 2 -6t! 1 *i 0 -985 1 '743 2 '583 2 -48 2 22 2 -07 0 -865 1 -37 . 9 ) . . . 5 *8i 7 '00 7 -39 8 -00 8.9 9 -0 8 -95 7 '15 7, 7 , .. i f 4 7 4'7 4 ' 2 2 -54 4 -15 7 -06 8 .6 12 9 6 10 -57 i'65 7 -42 7 -29 6 -7 6)'23 4 -95 .. - 72 H,O LisO B e 0 BsO, C 0 - N. Oj XgO 810. P.O. so. K20 CaO Sc. 0. R' a. 0 Al. 0. .. T10. .. v. 0 5 Cr03 3fIl.O. Fe.0. coo Co. 0. N i 0 pu'i203 CU. 0 zdb .. GeO. A+Oj Y Y SeO. SrO ZrO. Nb2O5 RuOs y 2 0 3 MOO. Ag. 0 CCiO 1.1.0. SnO. Sb. Oj Tdb. 1.0. .. E . .. Densities ... 0.917 2 -108 3 -086 1 7 9 1.457 1 -64 2 -805 3 -64 3 *94 2 -65 2 -22 2 -38 1 -936 2 -636 3 -25 3 -8 4 -25 4 -56 3 -35 2 -8 4 -75 5 '3 5 -6 4.8 5 -6 4 '8 6 -13 5 -75 5 '5 5 '73 4 '7 3 . '7 4-2 3'954 4 -5.1 5 -03 5 '85 4 '46 3 -92 7 '4 7 5 2 8 -2 8 -11 7 -18 6 - 7 3 i 8 6 -52 5 -1 4 -25 5 -02 F . Volume atoms of 3lenients . 0 *6149 0 -0806 0 -2029 0 -2429 0 '14 16 0 -0419 0 *Oil6 0 -0955 0 *0875 0 -0714 0 -0645 0 -022L 0 *0390 .- -.- . . . .I 0 -1141 0 -1 168 0 -1342 0 -1484 0.1515 0 *1534 0 -1367 0 -1094 0 -0i55 0 -3684 0 -0531 0 -029 0 *0461 0.0751 0 *0897 0 -1217 0 * 0987 0 -0771 0 -0651 0 * O f i l l 0 -0556 0 -049i 0 -0393 . . . . . . . . . G . ..- EiB . ..- 0 *9170 0 -2994 0 -3392 0 -1621 0 -1214 0 -1 168 0 -1217 0 -1498 0 -1458 0 -0935 0 -0.783 0 -0767 0 -0603 0 -0678 0-08 LO 0 -0561 0 -0883 0 -0947 0 -0651 0 4535 0 -0b63 0 -0946 0 -0954 0 -0817 0 *0955 0 -0819 0 -0966 0 * o m 0 -0842 0 -0877 0 -0650 0 -0492 0 -0559 0 -0500 0 -0515 0 -0565 0 -0650 0 ~04.74 0 -0409 0 -0728 0 50696 0 -0759 0 -05 23 0 -0631 0 -0562 0 9514 0 -05.21 0 -0408 0 -0335 0 -0395114 OXIDES OF THE ELEMENTS AND THE PERIODIC: LAW. , , ...... XTra-nium.. .... A. -- 2id S O Elements. ~~ Barium. ...... Lanthanum.. .. Cerium .......Tsntalurn ..... Tungsten.. .... Mercury ...... Lead ......... , . . . . . . . . ,, .... . , ....... , , ..... B. -- Atomic weights. 137 -0 13!! ‘6 -- lib: -3 lii *o , I 182 -5 200 *2 206 -93 TABLE 111-continued. C. -- Densities. D. Oxides. BaO La,03 3 , C& wb3 9 , Ta,O, HgO Bi203 Th02 P b 0 2 U S E. --- Densities. 4 -0 5 . 7 2 5 -9s 6 53 6 -0 6 -93 i -35 8 -01 7.32 11 -34 8 -89 8 -08 9 -S6 5 -26 10 *2 F. Volume atoms of elements. 0 -0291 0 *044S 0 -0456 9 -0574 ir -1048 0 9709 0 ‘0542 0.0469 0 -0476 0 9779 --- --- - - - - - 0 *0291 0.0417 0 -0428 0 -04‘7 1 0 -0427 0 -04993 @ *om2 0 ’0438 0 -0397 0 -0566 0 -04.29 0 -0388 0 -0424 0 ‘0438 0 ‘02 19 On Diagram 3, this is repented, fol- we have (Rb ?), SrO, YzOs, and ZrO, on a line, and then a decrease in the values of the ordinates of Nb,O, and Moo,.Tantalum and tungsten also show a similar change. In all three series the change first shows itself in elements of the fifth class of Table I. Owing to the high density ascribed to tungsten, the volume heat and volume atom of that element did not agree, OIL the volume heat diagram, with those of osmium and tantalam. From Table 111, it mill be seen that the density of the oxide is not abnoi.ma1. Where, on the diagrams, the ordinates are denoted by circles, the ordinates are not consecutive ones, and where halfmoons are given, the values have not been determined. By placing another element to fill in the gap between hydrogen and lithium, boron becomes the fifth element of ihe first series; indeed it becomes the analogue of vanadium. niobium, and tantalum. The same decreasein the value of the ordinate then occurs in all four cases (with coosecutive oxides) to the fifth element of the odd series, and although the chen;ical properties, and therefore the class oxides, of the elements of the first series are not in agreement with those of the third, fifth, and ninth, the fact that this change of density is common to all the series appears to give considerable support to the f andamental periodic basis indicated in Table I. OF course this classification somewhat alters the basis upon which MendeGeff predicted some new elements, for lie makes scandiumAVAILABLE “MINERAL” PLANT FOOD IN SOILS. 115 ekaboron, because it forms an oxide, SGO,, and becaase it occurs in the Same group as boron in his table. Fundamentally scandium is probably ekalithinm, and vanadium probably ekaboron. Similarly copper is ekasodinm, and zinc is ekamagnesinm. In column F, Table III, I bave given the volume atoms, so that they may be compared with the values in Column G. In the even series, there is, on the whole, a fairly concurrent change in the two values.

ISSN:0368-1645    

DOI:10.1039/CT8946500106

出版商:RSC    

年代:1894

数据来源: RSC

摘要:

AVAILABLE “MINERAL,’ PLANT FOOD IN SOILS. 115 XV.-On the Analytical Determination of probably ccuailable ‘‘ iiineral ” Plant Food iia Soils. (Illustrated by Examination of the Permanent Barley Soil of Hoos Fie1 d, Rothamsted. ) By BERNARD DYER, D.Sc. THE chemical analysis of soils, which in the early days of agricul- tural chemistry was looked upon as likely to be of very great prac- tical use in agriculture, mas soon found to be, as ordinarily practised, of very limited value. Determinations in the soil of the total quantities of the more important mineral elements of plant food have been long recognised as affording useful information only in exceptional cases: and even in these exceptional cases the results obtained have rather afforded “ probable indications ” than absolute information.Thus the fact that a soil contains much less phosphoric acid than is contained in average soils is a “ probable indication ” that it is in need of phosphatic manure; and the fact that another soil is much poorer in potash than average soils is regarded as a “ probable indication” that it needs potassic manure ; and in ex- treme cases these indications, read by experienced interpreters, may assume a degree of probability so great as to become all but certain, On the other hand, there are soils in which the proportions of total phosphoric acid or of potash are so large as to leave an experienced agricultural chemist no difficulty in forming an opinion that special applications to them of phosphates or of potash would be superfluous. But in the great majority of cases a soil analysis, as usually carried out, leaves us really in the dark, except as to such broad (and ad- mittedly valuable) general facts as richness or poverty in lime, pre- ponderance of sand or clay, or peat, h.The reason, as has often been pointed out, is that an analysis of soil, as ordinarily made, shows the total percentage of its coi~stitnents, VOL. LXV. X116 DYER : AVAILABLE ‘‘ 35NERAL ” or, at any rate, the percentage dissolved by strong mineral acids, witohout reference to the fact that only CL very small proportion of this total may be .available for plant use. For example, it is very usual to find about 0.15 per cent. of phos- phoric acid i n an ordinary average English agricultural soil. An average loamy soil, 9 inches deep, in its dry state, may, roughly, be said to weigh 1200 to 1500 tons per acre.Such a soil, contain- ing 0.15 per cent. of phosphoric acid, would accordingly contain somewhere about 2 tons of phosphoric acid to the acre, disregarding the subsoil altogether. Such a soil contains as much phosphoric acid per acre as would be contained in about 17 tons of superphosphate, or in nearly 10 tons of bone-dust : and yet on such a soil the addition of a mere few hundredweights of phosphatic manure may make the difference between a full crop of turnips and a bad one ! And similar statements would apply to other constituents of the soil. The obvious and thoroughly well recognised explanation of such anomalies is that i t is not the total proportion of phosphoric acid, or of potash, or of nitrogen, that rules a soil’s fertility, but the propor- tion of each of them that is present in an immediately available con- dition.It is true that the presence of a large totu2 of any constituent renders it, under many circumstances, prohable that there is present n larger available qnantity than is likely to exist where the total is small ; but this merely brings us to the “ probable indications ” which have been referred to as generally all that, are afforded by the ordinary soil analysis. By far the great’er part of the work done of late years in soil chemistry has had reference more especially to nitrogen, and quite an army of brilliant workers, at home and abroad, have engaged themselves on this branch of the subject. Much less general attention has, however, been given to the mineral chemistry of the soil, no doubt for the reason that, both scientifically and economically, it is less interesting, and that its unsolved problems have lacked the magnitude of those which till lately were involved, and indeed may still be said t o be involred, in the chemistry of nitrogen assimilation.The desirability, however, of adopting, if possible, some mode of analysis that should distinguish between total mineral constituents and those small portions of them probably available as plant food has not only been recognised, but has formed the subject of many suggestions and of some investigation. Closely allied t o this ques- tion is that of the “ available ” as distinguished from “ total ” ingrediel;ts in manures, and among those who have specially worked in the laboratory on one or both of these allied subjects from one point of view cr another will be fonnd the names of Berthelot,PIANT FOOD IN SOILS.117 DehBrain, Dngast, Eggertz, Fleischer, FreseEius, Grape, Gladding, HiIgard, Joulie, Lechartier, H. von Liebig, Lloyd, Luck, Maercker, Mallet, Neubauer, Nilson, Ollech, Petermann, Karl Schmidt, Stntzer, Thomson, Tollens, Voelcker, Vogel, P. Wagner, R. Wagner, Way, Wiklund, &c.; whilst a vast storehouse of data of information for future use has been accumulated, and is being yearly accumulated, in the field experiments of Lawes and Gilbert, at Rothamsted, followed up by those of late years carried on at Wobnrn by A. Voelcker and J. A. Voelcker. Most of the memoirs touching these subjects published by the various authors have: however, had to do rather with manures than directly with soils, and for the moment it is proposed to select for COR- sideration such work as has borne more directly on the question of soil analysis, so far as concerns phosphoric acid and potash.The memoirs coming strictly under this head to be found in modern chemical literature are comparatively few. The earliest paper of any importance that the author has been able to find is that of 3. v. Liebig (Zeitschrift d. Landwirth. Vereines, 1872), who worked on soils obtained from the permanent wheat field at Rothamsted. The attention of the present author was directed to Liebig’s investigation by Dr. (now Sir J. Henry) Gilbert, when some three or four years ago he discussed with him the lines on which it was proposed to pursue the investigation which forms the principal subject of the present paper.In the light of the results of the ex- periments since made with soils from other Rothamsted plots, plots not only more numerous, but experimentally nearly twenty yeare older, than those worked on by Liebig in 1872, it appears to be desirable to quote here some of Liebig’s figures. The specimens of soil used by Liebig were from Broadbalk WheaG field, Rothamsted, and represented the 1st and 2nd depths of 9 inches each. No. 3 soil had been continuously unmanured for nearly thirty years. Liebig determined in it several of the mineral constituents solnble in L‘ dilute hydrochloric acid’’ and in “ dilute acetic acid.” As the present paper is concerned only with potash and phosphoric acid, it is unnecessary to quote Liebig’s determinations of lime, magnesb, soda, &c.The other soils were 10 A, which for nearly 30 years had received an annual dressing of 400 lbs. ammonia salts, without mineral manure ; 5 A, which for a like time had received a complete annual dressing of mineral manures, without nitrogen ; 7 A, which had re- ceived annually both 400 lbs. ammonia salts and a complete mineral manure; and 2, which had received an annual dressing of 14 t ~ n s of farm-yard manure. K 2118 DYER : AVAILABLE ‘‘ MINERAL ” I I , Soluble i Total. in hot 10 per I cent. HC1. ------ - Potash.. .............. 1 i -970 0 -562 Phosphoric acid.. ...... i 0 *I71 .. The following are his results :- ~ Soluble in cold 5 per cent.HCl. 0 -090 0 -051 No. 3. Continuously unmanured. 1st 9 inches .................. 2nd 9 inches.. ................ 1st 9 inches .................. 2nd 9 inches.. ................ No. 5 A. Mixed mineral manure 1st 9 inches .................. 2nd 9 inches.. ................ No. 7 A. Mixed minerals and am- 1st 9 inches .................. 2nd 9 inches.. ................ 1st 9 inches .................. 2nd 9 inches.. ................ No. 10 A. Ammonia ealts only. without nitrogen. monia salts. No. 2. Farm-yard manure. Potash soluble in “ dilute hydrochloric acid.” 0 -085 Cot determined 7 7 7 7 7 7 7 ) 7 7 7 7 7 ) 9 ) 7 7 7 7 7 ) f 3 Potash soluble in ‘‘ dilute acetic acid.” 0 -015 0 *018 0 ‘013 0 -019 0 ‘038 0.022 0 ‘039 0 -018 0 -041 0 -0% Phosphoric acid soluble in “dilute nitric acid.” 0 -075 0 -047 0 ‘076 0 *047 0 -108 0 -058 0 *126 0 *061 0 -093 0 -065 These results afford indications that the investigation might have presented considerable interest had it been at the time systematically worked out.As it stands, however, its interest is greatlydiminished owing to the facts that no information is given about the degree of dilution of any of the acids used; and that only in one case is any comparison afforded between the action of the two acids-of unknown dilution-used to extract the potash. In 1880, Professor Karl Schmidt, of Dorpat, published a series of analyses of eight Russian soils, determining the total potash and phosphoric acid and also the percentages of these constituents soluble in hot 10 per cent.hydrochloric acid, cold 5 per cent’. hydrochloric acid, and cold 1 per cent. hydrochloric acid. The following average results of his determinations on three soils will illustrate his general results :- Soluble in cold 1 per cent,. HC1. 0 *084 0 ‘049PLANT FOOD IN SOILS. 11 9 Aqua regia. Ammonium Ammonium oxalate. citrate. 0 0 -056 -048 0 0 -042 -034 In 1881, P. P. Dehhrain observes (Ann. Agron., 6,392-3933 that in the neighbourhood of Grignon phosphatic manures are found of no use. The soil, he says, does not contain mom than an average quantity of phosphoric acid, but horn one half to one quarter of the total phosphoric acid present is soluble in acetic acid. He thinks it may be true generally that soils which contain phosphoric acid removable by acetic acid will not be benefited by phosphatic manures.Leaving chronological order for the moment, it is interesting to record that Dehhrain has recently published--eleven years after the observations just cited-a further paper on this subject (Ann. Agrm., 17, 4A5-454). In plots, at Grignon, unmanured since l87F1, he still finds 0-1 per cent. of phosphoric acid; yet on snch plots a moderate dressing of phosphatic manure now suffices to nearly treble the yield of wheat, showing, in accordance with general ob- servation and experience, that a soil fairly rich in its total mineral contents may nevertheless be exceedingly poor in assirnilab Ze mineral food. Pursuing the suggestion given in his earlier paper, viz., to use acetic acid as a means of diagnosing between assimilable and non- assimilable phosphates, he finds that, while the plots manured with phosphatic manures yield appreciable quantities of phosphoric acid to the action of acetic acid, the phosphate-exhausted soils yield only in- significant quantities.Plots that in 1879 contained as much as 0.03 per cent. of P,05 soluble in acetic acid, and that have been con- tinuonsiy unmanured, now yield but traces. In 1882, A. Vogel (Bied. Centr., p. 852) suggests that if a sample of soil tested with acetic acid yields no indications of phosphoric acid, its percentage of phosphates should be regarded as abnormally low. In 1884, Dngast ( A m . Agron., 9, 47-78) publishes the results of an elaborate investigation into the composition of three soils (from March6 Nenf, Loire Inf6rieure), A, B, and C, which, in their crop- ping, show different degrees of fertility.The only points that mljd here be quoted are the comparative proportions of phosphoric acid extracted from A and B by various reagents as follows:- ' water 1 with C02. wakw* Acetic acid. 1 saturated Distilled O ' O U --;Ti 0-013 0'013 0.015 0 moll120 DYER : AVAILABLE ‘’ MINERAL ” There appears to be not very much here to lay hold of beyond the fsct that the phosphoric acid exists in various conditions of solubility. (The phosphoric acid solnble in water is so enormoudy dispropor- tionate to the water-soluble phosphoric acid found in fertile soils by the present author, that these numbers appear to him scarcely credible, unless perhaps in the case of soils containing an enormous quantity of vegetable remains, which does not appear to havc been the case here.) In the same year (1884), G.Lechartier (Compt. rend., 98, 1058 -1061) suggests that a 2 per cent. solution of ammonium oxalate may be used as a means of determining t,he degree of solubility of the fertilising constituents of a soil. In 1889, Eggertz and Nilson (Bied. Centr., 1889, 664-668) pub- lished experiments indicating that extraction of a soil with hydro- chloric acid of 2 per cent. strength so far removed available plant food that bade7 could not grow in the washed soil, while extraction with 4 per cent. acid rendered it sterile to oats. In 1892 (Land,w. Jahrb., 20, 909-928), Wiklund published a work in the same direction, and expresses dissatisfaction with the method of Eggertz and Nilson as one for general adoption.He deals, how- ever, mainly with the distinction between organic and inorganic: phosphorus and sulphur in the soil. Lately, Berthelot and Andre have worke.3 a good deal (see various papers in Conzptes rendus) at the subject of the condition of the mineral ingredients of the soil, but apparentlr with more negative than positive results, so far as regards any conclusion as to the best mode of distinguishing between available and non-available plant food. The classical experiments of Way and A. Voelcker on the action of various saline solutions on soils, which are to be found in the back volumes of the Journal of the Royal Agricultural Society, are familiar to most agricultural chemists. They dealt rather with the power of soils to retain various manurial substances added to them, and with the liberation of locked-up plant food by the action of such indirect manures as soda salts, than with means for distinguishing between ‘‘ available ” and ‘‘ latent ” mineral “ food.” The literature of agricultural chemistry has become a wide one, and there are no doubt in existence many other contributions to this particular branch of soil chemistry.The papers that have been re- femed to, however, include all of any considerable importance that the present author has been able t o find. Many papers, however, as has been already said, exist on the kindred question of the assimila- bility of manures, which wil be referred to in a later section of thisPLANT FOOD IN SOILS.121 paper, the attention of the reader being, for the present, direcfed rather to the soil. It may be said to have been pretty widely recognised that some very much weaker solvent than strong mineral acid ought to be used in soil analyses, if these are to be of much use as indications of the proportion of available mineral plant food. Most of the suggestions made have, however, been arbitrary in the sense of not having any definite, or at any rate not, any defined, basis beyond the recognised necessity that the solvent should be weak-weak, that is to say, com- pared with concentrated mineral acids. For the couimercial distinction between phosphates presumed to be presently available and those presumed to be not presently avail- able in manures, it has for some years been customary in some .Coati- nentd countries, and also in America, to use as a solvent an ammonium citiate solution, originally, no doubt, because " precipitated " phos- phate, known from experience to be a readily available plant food, is soluble in such a solution. But much discussion has been spent on the point, and great diversity of views expressed, although in certain countries agreements have been come to for the arbitrary use for such purposes of citra te solutions of prescribed strengths and degrees of acidity or alkalinity.In 1884, Stutzer (Chem. I d , 1864, Feb.) raised an emphatic protest against the, by that time, widely accepted use of a strong solution of ammonium citrate as a means of estimating available phosphoric acid in manures, while granting that it might well serve to det'ermine " pre- cipitated " phosphate.He accordingly experimented with citric acid. which had been previously suggested by Tollens (Journa2 fur Land- zoirt., 30. Bd.)" on various phosphatic materials in use as manures. The details of Stutzer's results need not here be given, since they deal not with the soil itself, but with manurial materials', the consideration of which for the moment is deferred. It suffices here to say that Stutzer's results showed that while there was in many cases but little coherence between the fertilising value of phosphatic materials as determined by practical experience and their solubility in ammonium citrate, there was, on the other hand, a much more satisfactory correspondence between their solubility in weak citric acid and their known manurial efficacy.The strength of the solution Stutzer employed in his experiments was 1 per cent., i.e., 10 grams of citric acid per litre of water. He suggested that such a solution might best be adopted as a test of the probable solubility of * This journal is not in the Library of the Chemical Society, and Tollens' paper is not reproduced in contemporary journals, so that the present author has not had the advantage of consulting the original.122 DYER : AVAILABLE “ MINERAL ” manures in the soil (BqenlGslichkeit). A. Thomson later (Chent. Ind., 1885) endorsed Stutzer’s suggestions. Neither Stntzer nor Thomson, however, appears to give any reason for the strength of citric mid solution adopted, beyond the fact that the results obtained with a solution of this strength (1 per cent.) showed a fair correspond- ence with the comparative efficacy accredited by practical experience t o the fertilisers examined.The general question of solvents as bearing on soil analysis was taken up in 1884 by the present author, and has from time to time been the subject of a good deal of intermittent work. Much of this waa tentative and disappointing, and its results not worth transcrib- ing. Some of the earlier experiments may, however, be recorded before paasing on to the more fystematic enquiry to which they led np. PRELIMINARY EXPERIMENTS. Among these earlier tentative experiments were determinations of solubility in various solvents of the phosphoric acid and potash con- tained in a clay soil from the neighbourhood of Tewkesbury.It was a soil about which the author had no very special information. It happened to be sent to him for advice or suggestions as to manurial treatment, with the information that it had been lately cropped with beam after wheat, following vetches heavily manured with dung and fed off with cake-fed sheep. Beyond these facts, and those ascertained by its analysis, the author had no information, and the soil was used for these tentative experiments merely because there happened to be a large bulk of the sample, and because it was a fair average specimen of a good heavy clay soil, with a good deal of reserve mineral food in it. On account of this very limited knowledge of its history and capabilities, any such interest as attaches to the soil experiments to be described later is not to be expected, but, as far as they go, the facts ascertained as to the condition of solubility of some of the constituents are chronicled for what they are worth.The dry soil subjected to a tolembly complete analysis was found to contain, per cent. :-PLANT FOOD IN SOILS. 123 ---I_-- Phosphoric acid.. ............. Potash.. ..................... Soda ........................ Lime ........................ Magnesia .................... Sulphuric acid.. .............. silica.. ............... Oxide of iron.. ........ Alumina .............. Lime ................. Potash. ............... soda ................. Phosphoric acid.. ...... Sulphuric acid ......... Magnesia ............. 0 *OOO6 0' 0046 o-oO60 0 -0302 0 *W - ~ - 4 -320 4 -749 0 -873 0 -860 0 -544 0 -138 0 -211 0.151 70.745 4 '993 11 -421 1 -363 1.1M 1 -812 0 -577 - - The solubility in water of the more important constituents was determined in each case by shaking up a weight of air-dried soil corresponding to 250 grams of dry soil with 1000 C.C. of water in a stoppered bottle, and allowing to stand with occasional agitation. At the conclusion of each experiment, an aliquot part of the liquid was filtered, and the clear liquid evaporated for analysis.Percentage (in the dry soil) of Constituents iEissolved by Distilled Water. Water - 4 Soil 1' ___ -c 1 In 2 days. In 8 days. 0 -0oO6 0 -0058 0 '0139 0 -0382 O-Oo40 0 -0288 In 28 days. 0 'oO06 0 -0059 0 '01 11 0 '0637 0 '6080 - The question afterwards arose as to how far the quantity of water used might be responsible for variations in experimentally indicated water soh bility.Fresh experiments therefore were made in the same way, but uing varying quantities of water.124 DYER : AVAILABLE "ERAL " Dissolved by Water in 8 days. Water - 10 Soil -i-' - Phosphoric a.cid.. ............. Potash.. ..................... Soda ........................ Lime ........................ Magnesia .................... 0 *0018 0 '0077 0 '0121 0 '1148 0 -0200 Evidently the qnantity of water used for a given weight of soil produces variations which, if there were any object in pursuing such a mode of soil esazination, would render it desirable to work under definite conditions. Probably the frequency of agitation would also bring about discrepancies (and there are some discrepancies here) when dealing with such small figures.Temperature too should have been regarded. As in each ca.se the experiments were made at the ordinary temperature of the laboratory, and as some months elapsed between the first and second set of water solubility determinations, some of t h e differences mag be due t o this. This point, however, is of little consequence, since mere water solubility is not an index of fertility. An atternRt was next made to determine the solubility of the vsfions constituents in carbonic acid water. Accordingly, a set of experi- ments mas made precisely like the last, except that the water used mas saturated with carbonic acid. Constituents Dissolced by a Solution of CO, in r a t e r .0 *OOOi Phosphoric acid. .............. I 0-0010 Potash.. ..................... 0-0067 0-0085 Soda ........................ 0' 0139 0'0150 Lime.. ....................... 0 -1204 0 '1 266 I Mttgnesia .................... i 0.0150 0 -0120 I n 28 dap. 0 -0009 0 'OOi3 0 -0087 0 -20T2 0.0300 Later, the following determinations were made of the quantity of phosphoric acid dissolved from this soil by solutions of ammonium citrate. The ammonium citmte was made from specially purified citric acid. In each case a quantity of air-dried soil equal t o 2550 grams of dry soil was shaken up with 1000 C.C. of water containing ammonium citrate, rendered alkaline by the addition of 25 C.C. of 0.880 ammonia,PLANT FOOD IN SOILS. 125 The liquid remained in contact with the soil for seven days (in late spring) with frequent shaking.Antmonium Citrate Ezperimmtd;. 1 Per cent. on the soil. Solrent 4 !-- --- Water only.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I Water and ammonia only.. .. . . . . . . . . . . . . . . . . . . . . . . . . . 0 '1 per cent. alkaline solution of ammonium citrate.. . . . . 7 Y Y I 0 '25 ...... 0 -5 .. .. .. 1 1 -0 .. .. .. I 2 - 5 .. .. .. ~ 5 ' 0 .. .... I 7 7 7 ......I ! I f 7 7 7 . 7 7 t 77 Y 7 7 7 I 10-0 15 '0 20 -0 30 -0 40'0 ...... i 50 -0 9 f 7 7 7 7 7 7 7 7 9 7 9 7 1 7 7 .. .. .. , .. .. .. j .. .. .. J . . . . . . , 0 -0oO9 0 -0034 0 -0036 0 -0042 0 -0046 0 *0056 0 -0080 0 *0090 0 '0131 0 -0161 0 -0154 0 -0194 0 -0220 0 -0260 1 0-0050 , 0.0047 ' 0.0059 0 -0079 0 -0139 0 -0156 0'0212 0 -0253 : 0-0268 \ 0-0261 : 0.0223 0 *0223 Some similar determinations of solubilities were also made in a clay loam soil from the neighbourhood of Chelmsford-the soil, indeed, of Mr.Rosling's farm, on which have been carried out for some years, with the co-operation of the author, the various series of field experiments made for the Essex Agricultural Society. This soil, on analysis, yielded the following results, the solvent, except for C1 and Nz05, being the usual one, viz., hydrochloric acid acting on the incinerated soil :- Fine Soil, Dried at 212" Fahr. Silica and silicates insoluble in hydrochloric acid.. Oxide of iron.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alumina.. . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . .. Lime ........................................ Magnesia . . . . . . . . .... . . . . . . . . . . . . . . . . . . . . . . . . . Potash . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . - . . . Soda.. ......... ............................ Carbonic acid.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phosphoric acid. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Snlphuric acid.. . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . Nitric acid . . . . . . . . . . . . . . . . . .... . . . . . . . . . . . . . . . Chlorine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . "Organic matter, water of combination, &c.. . . . . . . 86.480 3.721 3.202 0.371 0-31 9 0.305 O-OiP 0.121 0.141 0.031 0.001 0-002 5-032 100~000 --- * Containing nitrogen, 0.176.126 DYER: AVAILABLE “MINERAL” The farm was one that had for some years been kept in good condi- tion by careful farming and liberal manuring, and the particular field mmpled showed itself able, in the year in which this sample was drawn, to produce some 16 tons of mangolds per acre, without the addition of any manure, after a preceding crop of oats.This shows it to have been pretty rich in available mineral food as well as iii nitrogen. Determinations of the phosphoric acid and potash, soluble in am- monium citrate solutions of various strengths, were made (as in the case of the soil already mentioned) with the following results :- Ammonium Citrate Experiments.Solvent - 4 Soil 1 ~ --. Time of contact 7 days. Water only .................................... Water and ammonia only.. ...................... 01 per cent. alkaline solution of ammonium citrate.. 025 9 9 1 9 0 5 Y Y 7 s 1.0 Y Y Y Y 2.5 J Y 99 5-0 9 9 Y Y 100 >> 9 9 15-0 9 ) Y9 200 Y9 n 30-0 Y 9 $ 9 4’0 Y Y Y9 50.0 >9 Y 9 Per cent. on the dry soil. Phosphoric acid. --- 0 * m 3 0 -0036 0 -0047 0 -004.6 0 -0o6O 0 -0084 0 -0128 0 -01443 0 -0149 0 -0163 0 -0207 0 -0232 0 -0237 0 -0248 Potash. 0 ‘0011 - - 0 -0037 0 -0077 0 aJ54 0 -0065 0 9093 0.0077 0.0082 0 *01M 0 -0097 0 -0110 - These results are, on the whole, not dissimilar from those obtained with the Tewkesbnry soil, except in potash, in which constituent the Tewkesbury soil was much richer. At about this time the author had become convinced, by experi- ments of his own on phosphatic manures (see later), that the conten- tion of Stutzer was right, and that the solvent action of an ammoniacal salt did not give a reliable index of availability.Weak organic acid had shown itself to give far more consistent and intelligible results with manures, and why not with soils, as suggested by H. v. Liebig, Dehhrain, and others already quoted ? But the first difficulty to be met-a very considerable one, as shown by the following experiments with the Chelmsford soil-was to decide the strength (or weakness) of the acid to be used.PLANT E'OOD IN SOILS. 127 Citric Acid Solution Experiments. Solvent - 44) Phosphoric acid --- soil 1 ' (per cent. on dry soil). 0.125 per cent. citric acid solution...... 0.0200 0.25 9 9 9 , ...... 0.0250 0.5 7, 7 , ...... 0.0264 1.0 7, 9 , ...... 0.0312 2.5 T Y $ 9 0.0728 5.0 ...... 0.0896 ...... 7 7 Y I Solvent 20 Soil I ~ = -. 0.25 per cent. citric acid solution. ...... 0.0197 Solrent 20 ~ . - - Soil - i* 0.5 per cent. citric acid solution.. ...... 0.0193 Solrent 10 Soil - 7 -- 1.0 per cent. citric acid solution.. ...... 0.0161 It is obvious that if citric acid be used the yariations are great, as. might be expected, in accordance with the degree of dilution and the. proportion of acid to soil, at any rate where the contact with the soil is of short duration, as in the case of these last enumerated experi- ments, in which only three days were allowed. Seeing that the end in view wa.s a means of distinguishing between matter that a plant can take up and matter that a plant cannot take up, the possibility suggested itself of arriving at some conclusion from a closer study of the means by which a plant collects its mineral food.The mere solvent action of water, or even of water saturated with carbonic acid, at all events of the very limited quantity of water that exists in even a wet soil, is altogether insnacient to account for the solution of all the mineral plant food required by a crop. N o doubt the action of decaying organic matter and of " hnmic " and " nlmic " acids, &c., produces some effect, but it seems probable that, useful as humus is in a soil for many purposes, the solvent action of its decomposition products on the minerals of the soil in actual fact has been overrated, and that the chief solvent agent for soil minerals is the root sap of the plant.EXPERIMEMTS ON THE ACIDITY OF XOOT SAP. It has long been accepted as a fact that plants help themselvee to a part of their mineral food by means of the solvent action of their128 DPER : ATAILAF3m ‘‘ MINERAL ” acid root sap on the particles of soil with which the rootlets come into contact. The observations most frequently quoted in support of this are the classical ones of Sachs, who showed that the acid of the root sap was sufficient to etch, by its corrosive action, the surface of polished marble buried beneath the soil. But the mere observation that roots have an acid sap is much older than any recognition of the significance of their acidity.For example, Philip Miller (gardener to the Society of Apothecaries) wrote in 1733 :-“When the juice enters the root, it is earthy, watery, Poor, and acid.” He also speaks of “that tart liquor oozing from the root of the walnut tree when cut off in the month of May.” The subject of the acidity of root sap and of its solvent action on certain soil constituents, but more especially on nitrogenous organic matter, is dealt with at some length in the paper “ On the Present Position of the Question of the Sources of Nitrogen of Vegetation,” pitblished i n 1889 in the Phil. Trans., by Sir John Lawes and Sir Henry Gilbert. I have since learned from Sir Henry Gilbert that, in the course of their investigation, the authors experimented on the acidity of the root sap of about 7-5 different plants. None of the quantitative determinations, however, have been published, and few, if any, other attempts appear to have been made to determine the degree of acidity of root sap.In order, therefore, to obtain some information on this point, the determinations that are now recorded were under- taken. The plants were, as far as possible, taken at a time when they were in vigorous growth. Graminaceous plants-cereals and grasses-were selected just before, or in the first stage of, inflorescence, and so also were leguminous plants. Potato plants were taken when the tubers were beginning to develop, mangolds and turnips when the bulbs had developed to a diameter of 1 to 2 inches, and cabbages and other crucifers at a corresponding stage of growth.In addition to ordinary agricultural plants, a large number of garden plants were examined-annuals, biennials, and perennials-all being taken, as far as was practicable, during active growth. The author desires here to tender his special thanks to Messrs. Sutton a9d Sons, Reading, for kindly placing at his disposal their pure cultivation plots of grasses and clovers, from which separate growths of the chief pasture grasses were obtained. The problem of ascertaining the acidity of the root sap is by no means a simple one. In the case of large, fleshy, tuberous, or bulbous roots, it is, of course, easy to express the juice. But the examination of such juice does not necessarily throw much light on the real question under investigation, because i t is the acidity of the sap, not of the main or tap roots, but of the small roots and rootlets, whichPLANT FOOD IN SOILS.129 determines the soil-solvent functions of the plant. It by no means follows that we know the acidity of the rootlet sap when we know the acidity of the juice of the fleshy tap root from which the rootlets spring. To mechanically express the sap from fine roots and rootlets is practically impossible. It is necessary, therefore, to proceed in some other fashion, and it is not easy to find one that is free from objection. The rootlets of most plants form a matted mass which clings intimately t o the soil, and, seeing that soil abounds in basic substances, it is necessary to free the rootlets from the adherent particles. The plan finally adopted was to dig up a large block of the soil in which the plant was growing (or, in the case of pot-grown plants, to take the whole potful of earth and roots), and to rapidly wash the earth off with a good flow of water over a bowl.After a little practice, it is not di6- cult to effect the washing very rapidly and without much bruising of the roots. After the removal of the earth, the mass of moist roots and rootlets is lightly disentangled and freed from external moisture by pads of blotting paper and by blowing on the roots with bellows. When the roots no longer moisten new blotting paper, they are taken as “dry.” In living and unbruised roots, the probability oE losing soluble constituents during the process of rapid washing is small, because transfusion through the walls of living cells takes place very slowly, and even the vessels whose open ends may be exposed to the water are of such capillary fineness that there is little likelihood of rapid diffusion taking place.At all events, no better method of obtaining the rootlets free from soil and in a condition likely to be little changed presented itself. A portion of the externally dry roots and rootlets is weighed and dried in the water oven. The moisture determined by loss in weight is taken as a sufficient approximation to the original “ moisture ” of the roots to furnish a basis for calculation of the sap. Another portion of the roots is quickly snipped up with scissors (not “ minced,” so as to expose any material portion of the sap to the action of the metal) and boiled in distilled water for some time.The liquor is then strained off, and the exhausted rootlets rubbed to a pulp in a porcelain mortar, and again bbiled. The total extract is boiled down to a small bulk to boil off carbonic acid, and the fixed acidity taken by titration with standard alkali, using phenolphthalein as an indicator. After a good many determinations had been made in this way, it was thought possible that a more rapid and satisfactory extrac- tion might be made if the dried rootlets could be used, as transfusion would be mcessnrily more easy. In a large number cjf cases, there- fore, the portion dried for the moisture determination was ground to130 DYER : AVAILABLE ‘‘ MINERAL ” powder in a mortar-a veryquick process in the case of the crisp dry roots-and the powder boiled with water, and the decoction concen- trated. It is not necessary to filter or pour it off, as neither tbe dark coloar nor the turbidity interferes with the reading when so strongly reacting an indicator as phenolphthalein is used.In most cases the determination of acidity in the dried roots agreed very well with that in the moist roots, and in future it would probably be adopted as more convenient. Of course it is open to the objection that some mutually destructive action may occur between organic acids and the various other organic substances present when dried together, but the drying occurs very rapidly, and the total quantity to be determined is so very small, that even a considerable per- centage error in the total acid present would little affect the percent- age on the sap.On the whole, although the whole process may be scientifically open to some objections, it appears sufficient to afford a fairly approximate idea of the percentage of total fixed acidity in the roots, and the moisture being known, the acidity f u r 100 parts of moisture mr-y be taken as giving some approximate figure for the “ sap acidity ” of the roots. No attempt has been made in any of these experiments t o identify the acid to which the acidity was due. Probably, a considerable variety of organic acids exist in the rootlets of a single species, and in different orders of plants different acids would predominate. For the special purpose the author had in riew, i t was convenient t o calculate the acidity into the form of citric acid, but in the record of results now given the acidity is also stated in terms of hydrogen.In each case will be found stated- i. The percentage of moisture in the root. ii. The percentage of acidity in terms of hydrogen. iii. The percentage of acidity in terms of citric acid. iv. The ratio of acidity to 100 parts of moisture (“ sap acidity ”) in v. The ratio of acidity to 100 parts of moisture (“ sap acidity ”) in terms of hydrogen. terms of citric acid.PLANT FOOD IN SOILS. 131132 DYER : AVAILABLE '' MINERAL "Oh. L 09.0 t9.0 m. 0 89.0 LrZo,o I L,T8 'I.8L GTZO. 0 k ......................... ( 6 ( 6 ......................... ( Lh .Ka 9 .M VJ'j s) 1') .l WflV.ti[ s9.90 ' 0 &BOO. 0 6600.0 Lfoo. 0 nL60. 0 4Z90.0 9980.0 09r7;O.0 9090 * 0 OCiLO. 0 ........................................ w a g 4.99 , - > V R D V S O ~ (iSbO.0 ROLO. 0 LLOO 0 li!?oo. (I 89.0 89.0 86.0 09.0 F9.0 82.0 09.0 z9. T €4. T 00.8 69. G 6900.0 8600.0 €€TO* 0 2400.0 4400.0 OPOO. 0 Z400*0 'I. 69 4 . ip9 4. 'F-8 0. €8 6. 88 R. 06 i p s €4 .................................. (( 1: ............................... ' '(&l)~D.MpaH 'W.aDVITV?Ig sil ................................ (hp3) zan.?,dr (qoam3) sno?av(jy (tT!rrs*InJ) woV,u:/?swJ ........................... (6 ......................... ............................... ........................... (A'apsCT) tunu?gaso.qaj * s ~ . a a a ~ ~ ~ x a ~ nDYER : AVAILABLE " MIPvTERAL " 0 5 3 0 0 3 00 C 0 O C 3 h h C O 0 0 0 0 0 0 0 0 u - 3 . . - . . . - . . . .. . . . . . . . . - . . . . . . . . . . . * . - . . . - . - . - . . . . . . . - . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - . . . . . . . . . . . . . . . a - . . . . . . - - . . . . . . . . n - - c) . . 24 - . . . . . 2 - 5 .- 2 : s : c l - 0 . . . . . . . . . . * . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I . . . . . . - . . . . . . . . . . . . . . . . . . . . . . . -PLANT FOOD IN SOILS. 135 3 0 0 0 3 3 0 0 0 0 0 0 0 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . n . . . . . . 1 5 : * * .n-C . . . . . . . a v - . . . . . I _ . . . . . . . . . . . . . . . . . . . . . I . ( . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . - . . . . . . . . . . . . . . . . . . . . . . . . . . . . . n . . % 2 : : 5 : : - . . U J136 DYER : AVAILABLE ‘‘ MINERAL ” 0 0 3 c 0 0 0 0 0 0 0 0 0 0 0 0 0 3 0 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . :nz:: : : : : : : : : : : : : : : : *cg-z . . - . . - . . . . . . . . * - & m o 2 . . . . . . . . . . . . . . . :,_a%? . . . . . . . . . . . . . . . : g g g s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .PLANT FOOD M SOILS. 137 There are recorded in the foregoing tables about 100 root acidity determinations, the plants being taken from 20 different natural orders. It will have been noticed that the variations are wide even in plants of the same order and even of the same species. The ayerage “ sap acidity ” of the roots of the 100 plants is- Expressed as hydrogen ............0.0122 Expressed as crystallised citric acid. , 0.8540 If we take, however, the average acidity of the plants in each order examined (since the number of plants examined in each order varied 1-ery much), and then average these averages, the average “sap acidity” for the roots of the 20 orders is- Expressed as hydrogen. ............. 0*01:3 Expressed as crystallised citric acid . . 0.910 Of course either mode of averaging is, for very obvious reasons, objectionable, as the manipulation of figures is often liable to be. The average figure-whichever mode of averaging be taken-diff ers widely from some of the individual figures. The average of the whole, however (0.91 per cent., reckoned as citric acid), represents very nearly the averages found in each case for the plants included in Ranun- czdacece, Crucifem, Ca?-yophyllacee, Legunzinosm, Onagracece, Araliacece, and Boraginacece, which averages vary between the limits of 0.81 per cent.and 1.12 per cent. Of the remaining orders, we have Tropceolacece, Primulacece, Umbel- Iqerw, Compositce, Campanulacece, Ckenopodincece, and Graminece, ranging from 0.53 per cent. to 0.68 per cent. DtjxsaceR and Solanacece (single species only) fall much below the average (0.44 per cent. and 0.34 per cent.), and so do the Liliacece (0.36 per cent.), though one of the only two species comprising the four liliaceous plants examined gave higher results. On the other hand, the plants examined in RosaceGe and Plumbaginece gave exceeding high results checked in each case by operating on two distinct plants of each species chosen.Obviously these determinations, numerous and laborious as they have been, can only be regarded as being in the nature of a tentative and preliminary enquiry of a very crude kind if criticised from the botanical or physiological standpoint. But they appear to be sufficient to indicate that the ratio of the soluble free acid in the roots of plants to the moisture contained in them-which is here called sap-acidity- probably generally falls within, and not very far within, 1 per cent., calculated as crptallised citric acid. Citric acid is chosen to express the acidity partly on account of its being an organic acid, and in that sense kindred to other root sap138 DYER : AVAJLABLE “ MINERAL ” acids ; partly because it is the acid generally used by those who have attempted to determine available phosphoric acid in manures by means of weak acid, in particular by Tollens, Stutzer, A.Thomson, and P. Wagner ; and partly because i t is at hand in every agricultural laborakory in a state of purity, and therefore a convenient acid. On the whole, these sap-acidity determinations, however desultory and imperfect in a scientific sense, seemed to confirm the wisdom of Stutzer in adopting (Tollens had suggest,ed various strengths) a 1 per cent. solution of citric acid as a standard test of the availability of phosphates in manures, though he appears to have lighted on that strength by experimental enquiry based on quite other grounds.ESPERIMESTS 0s ROTHANSTED EARIiET SOILS. Since the choice of a solvent for use in soil analyses must in the end be empirical, both as regards its form and its strength, it seemed now worth while to make some ef‘fort to t,est the expediency of adopt- ing a 1 per cent. citric acid solution. Such a solution appears to yield instructire information in the case of manures, and it approxi- mates fairly closely t o the average acidity of root sap. It remained, however, to see how far its solvent action on the constituents of the soil might afford a real measure of the soil’s present fertility so far as such fertility is affected by the quantity present of available mineral elements of plant food. In order to do this, it was necessary, clearly, to operate on soils about whose history and whose fertility very exact information was attainable, and the one spot to which an agricultural chemist in England or elsewhere would naturally turn for such a purpose at once suggested itself as a source of material whereon to work.By the kind permission of Sir John Lawes, and with the advice and personal assistance of Sir Henry Gilbert, the author was fortunate in being allowed t o draw a complete set of samples of soil from the world-famed Hoos Field at Rothamsted, on which barley has been grown for 40 years in succession, and on which each plot has been year after year subjected to some one kind of unvarying manurial treatment. A precise record having been preserved, not only of the manures applied to each plot, but also of its yield of grain and straw year by year, a study of a summary of the field’s history is sufficient to show which plots are langnishirig for phos- phoric acid, which for potash, which for nitrogen, and which, in rarjing degree, for all.After so many years of treatment, the knowledge of the manures used almost suffices to predict the crop of an average season; while the yield of crop would, on the other hand, almost suffice to indicate how i t has been manured-so closely do the two factors-the cc priori one of treatment and the a posterioriPLANT FOOD IS SO- 139 one of yield-independently point, when properly studied, to the same degree of fertility. Any mode of soil analysis that would furnish a third scale of indications pointing to the sawe conclusions might almost be accepted as a prmisionally satisfactory method of soil analysis-regarding the mineral elements of fertility, with which alone we are at present concerned. Whatever might be the result of the proposed experiments, it could scarcely be doubted that they would bring to light some points of interest-an anticipation that mill be seen to hare been fully realised.The samples were drawn in the autumn of 1889. A square sampling iron, such as has been described in papers from Rothamsted, was used. I t is so made that it drams a fair sample of the soil to a depth of 9 inches. To sample each plot, samples were drawn in this may from several places and mixed together on a sheet. 20 lbs. of the moist earth, including stones and pebbles, were then weighed out from the mixture, bagged, and sent to the author's laboratory in Loudon.22 out of the total 29 plots in Hoos Field were thus sampled. The sample in each case was allowed to become sn5ciently air- dried to allow of its being crumbled and sifted. The stones-flints, pebbles, &.-that would not pass through a $inch sieve were re- jected, and the moisture in the air-dried soil was determined by drying a fair sample of it at 100" C. The analytical determinations of total phosphoric acid and potash mere made on the dry soil. For all the solubility experiments, however, the &-dried soil was used, a quantity being taken which ic each case corresponded t o a given weight of actually dry soil ; for it was considered that to completely dry the soil before weighing it out for the solubility experiments might be t o risk, by dehydration, the production of some change in the composition or constitution of the phosphoric acid or potash com- pounds existing in the soil, and possibly therefrom the bringing about of some artificial modification of their solubility.The percentage resclts obtained have been also calculated out as pounds per acre. In these calculations the author has availed himself <if data kindlysupplied to him by Sir Henry Gilbert, as to the acreage weights of fine soil on the experimental plots. These are- Weight per acre of dry soil (excluding stones retained by $-inch sieve). First 9 inches foil Seiies 0, A, AL4, and AAS (mear, of 16 plots and 56 samples) ...... 2,527,879 lbs. For Series C (mean of 4 plots and 16 samples) ............................ 2,361,461 lbs.For 7' (mean of 3 samples) ............... 2,486,870 lbs. 2,064,567 lbs. For 72 (mean of 3 samples). ..............140 DYER : AVAILABLE " MINERAL " Perhaps a few words may be said as to the analytical methods adopted. The total phosphoric acid was determined in each case in 10 grams of the dried soil and also in 25 grams, the mean of the two results being taken. The numbers obtained in each case were, however, all but identical, the difference in the duplicate percentages being in most cases only a small one in the third place of decimals. The soil was incinerated and digested with hydrochloric acid, and evaporated to dryness, redigested with acid, filtered, and washed. The filtrate and washings were concentrated to a small bulk, and treated, in the cold, with excess of a solution of ammonium molybdate in nitric acid. After standing 48 hours, the liquor was decanted through a filter, the precipitate washed several times by decantation, first with dilute acid, then with pure water in very small doses, and fii~ally transferred to the filter and washed free from excess of acid- The ammonium phosphomolybdate was then dissolved in ammonia, evaporated to dryness in a platinum capsule, and dried to constancy in a, water-oven.The residue contains 3.5 per cent. of its weight of phosphoric acid. This is the method of 0. Hehner (AnaZysf, 1879), and for determining small quantities of phosphoric acid such as occur in soils or in solutions of iron and steel, is very mnch to be preferred to the old-fashioned method of conversion into magnesium-ammonium phosphate.The solubility of the yellow precipitate in the small quantity of wash water used is in most cases negligible. As A matter of fact, the quantity of mash water used in these analyses was found capable of dissolving only 0.005 gram of precipitate, of which only 0.00017 is phosphoric acid, making an error of 0.0017 per cent. on the soil if 10 grams be used, or of only 0.0006 if 25 grams be used. I n the citric acid experiments, t(o be presently described, the solution from 50 grams of soil was used, when the error due to solubility of precipitate shrinks to 0-0003 per cent. The correction for this solu- bility was, however, made in each case. It may be observed that the method of Hehner is not applicable if the molybdic solution be added to a hot liquid, since, in that case, some molybdic acid is sure to crystallise out with the yellow preci- pitate.Moderate and careful warming to about 35O C. hastens pre- cipitation, but i t is preferable, when speed is not a special object, to precipitate cold, and leave the beaker standing at the laboratory temperature over night, or longer if the quantity to be determined is very minute. To determine potash, 10 grams of fine, dry soil were treated with 10 C.C. of hydrochloric acid and evaporated to dryness on the water- bath, the residue taken up with another 10 C.C. of acid, warmed, diluted with water, boiled, filtered, and washed. The filtrate andPLAXT FOOD IS SOILS. 141 washings were concentrated and gently incinerated to get rid of organic matter, and the residue redissolved in hydrochloric acid, and evaporated slowly down with a considerable quantity of platinum chloride.If the evaporation be conducted slowly, the platinum- potassium chloride settles out well, despite the iron, aluminium, and calcium salts, and is easilF washed with some more platinum chloride solution, followed by alcohol. The application of this (Tatlock’s) modification of the platinum chloride process to solutions containing comparatively minute quantities of potash amid an overwhelming excess of iron, aluminium, and calcium salts is probably new to many chemists. It works admirably, and obviates the necessity for re- moving iron, aluminium, calcium, magnesium, &c., with the necessary use of ammonia, and the tedious processes of concentration and final 1-olatilisation of the ammonium salts; but, of course, the process cannot be employed if soda also is to be determined.It should be observed that the filtrates from several determinations were collected and very exhaustively examined for any possible trace of unpreci- pitated potash, but none could be detected. The potash soluble in hydrochloric acid having been thus deter- mined, the undissolved siliceous matter was incinerated, weighed, and finely ground in an agate mortar. A weighed portion of it was then mixed with a large bulk of pure calcium carbonate and a little ammonium chloride and heated, beginning with a low temperature, rising slowly to bright redness. The mass was then boiled with water, washed, incinerated, re-ground, mixed with some more am- monium chloride, and again heated, boiled, and mashed out.The process was repeated once again, and the filtrates from all the treat- ments coccentrated, the calcium being removed as carbonate, and the potash determined in the filtrate, after evaporation and incineration at a low temperature, by means of platinum chloride. It was found, in repetitiou analyses, impossible to get constant figures for potash soluble in hydrochloric acid, for even a difference in the time taken to evaporate to dryness with acid was found to affect the result. But all the figures stated under this head were obtained by vwrking under similar conditions, and they are such as are obtained in ordinary soil analysis. The determinations of solubility in citric acid solution were made as follows :-A weight of air-dried soil corresponding to 200 grams of completely dry soil was placed in a Winchester quart bottle with 2 litres of distilled water in which were dissolvcd 20 grams of pure citric acid.(Winchester quarts which had been used for the storage of strong acids were chosen in order to eliminate the possibility of dissolving potash, &c., from the glass. The bottles were rinsed many times, allowed t o rest full of water for some days, and again rinsed.)142 DYER : AVAILABLE '' MINERAL " The soil was allowed to remain in contact with the 1 per cent. citric acid solution for seven days, being, except on one day, shaken up a great many times each day, whenever, that is to say, the soil had settled well down.The bottles rested, however, at night. Altogether the shaking up, which war not violent, but only sufficient to break up the cake of mud which formed each time the soil settled down, may have been performed on each sample about 400 times. The temperature ranged between 10" and 19" C., and averaged about 15" C. The es- periments being made in winter, it mas practically impossible to keep a constant temperature in the laborator7, and the bottles 3ver.e too large and too numerous to be placed in any ordinary constant temperature closet. On the whole, the conditions throughout the 22 experiments may be said to have been fairly even. After seven days' treatment as above, the solutions were filtered ; 500 c.c., i.e., the solution from 50 grams of soil, mere used for each determination.I n each case the solution was evaporated to dryness in a platinum basin, and gently incinerated at a low temperature. The residue was dissolved iu pure hydrochloric acid, evaporated to dryness, redissolved, and filtered. In the filtrate the phosphoric acid or potash, as the case might be, was determined by the methods already described, and, as the actual precipitate weighed represented 50 grams of soil, the multiplication of any experimental error in con- verting into percentages was minimised. The methods of sampling, of analysis, and of experiment, having been sufficiently described, the results may now be recorded and discussed. For the better study of these, however, it is necessary to prefix a descriptive list of the plots of soil represented, showing fully the manurial treatment that each had received during the 38 years of continuous barley growing, the average yield in grain and straw for the first half of that period, for the second half and for the whole period, and the yield in the year (1890) following that in which the soils.mere sampled. These data are transcribed from the annual state- ment of experimental results issued from Rothamsted.PLANT FOOD IN SOITS. 143 c c E c I - z G c e : : 5- :5 : : 2 : : : 2 . . . .= : :"o :.5" : : :s : $ : 000 0114 DYER : AVAILABLE '' 31INERAL "PLANT FOOD IN SOILS. 145 The following tables show the results of the determinations of total phosphoric acid, and also of the porttion of phosphoric acid dis- solved by the 1 per cent.solution of citric acid, expressed as per- centages of the “fiue,” dry soil, and also as (estimated) lbs. per acre :- Phosphoric Acid Determinations in Samples cf Badey Soils from HOOS Field, Rothamstea. Jlanure applied erery gear since 1852 (for quantities see pages 143 and 14-4). -I_------- 1 0. S o manure .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. 2 0. Superphosphate.. .. .. .. .. .. .. .. . . .. .. . .. .. 3 0. Potash, &c. (no phosphates) . . .. . . . . . . . . . . . . 4 0. SuperphoJphate, patash, 8rc. .. .. .. . . .. . . . . . . 1 A. Ammonia salts.. .. . . . . . . . . .. . . .. .. .. . . . . . . 3 A. Ditto and superphosphate.. . . . . . , . . . . . . . . . 3 A. Ditto, and potash, &c. (no phosphate) . . . . .. . 4 A.Ditto, superphosphate, and potash, 6rc ... . . . . . 1 AA. Kitrate of soda.. . . . . .. . . . . . . . . . . . . . . . . . . 3 AA. Ditto and superphvsphate . . . . . . . . . . . . . . . . 3 AA. Ditto and potash, &c. (no phosphates). . . . ,- 4 A& Ditto, superphosphate, and potash, 8-c. .. .. I 1 AAS. Sitrate of soda and silicate of soda.. . . . . . . 2 U S . Ditto, ditto, and superphosphate . . . . . . . . , 3 4 A S . Ditto, ditto, and potash, &c. (no rho-ll - I iates 4 188. Ditto, ditto, superphosphate, and potash, 6rc 1 C. Rape-cake.. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. 2 C. Ditto and superphosphate.. .. . . . . . . . . .. . . . . 3 C. Ditto and potash, &c. (no phosphates) . . . . .. I 4-C. Ditto, superphosphate and potash, kc.. . . .. . I 5’. Farm.yard manure for 20 years, iininanured for last 18 Fears.. . . . . .. . . . . . . . . . . . . . . . . . . . . . . 72. Farm-yard manure for 38 Sears.. . . . . . . . . . . . . Percentage of P205 in fine soil, calculated on dry state. 1 P20, dissolved 0 -099 0 -182 0 ‘121 0 -189 0 -097 0.173 0’105 0 -182 0 -105 0 -165 0 ‘104 0.179 0 -106 0 -180 0 -105 0 -169 0 -158 0 -229 0 -155 0 -203 0 -134 0 -1i6 1 I 1 I I I i i I I I 0 -0055 0 *0@3 0~0400 0 -053g 0 -0060 0 * 0425 0.00~1 0 ‘0500 0 ’0067 0 *0350 0‘0068 0.0475 0- 0074 o ,047s 0 - 0 4 4 2 0.0479 0.01&7 o -0214 0 -0636 0 .&-63 0 -0206 0 *0447146 DYER : AVAILABLE ‘‘ NINERAL ” Phosphwic Acid. Percentage results calculated into lbs. per acre, on the assumption that t!ie tlmples drawn fairly represented the soil to a depth of 9 inches. Manure applied every -year since 1853 (for quantities see pages 143 and la). 1 0.No manure .................... 2 0. Superphosphate.. ............... 3 0. Pot,ash, &c. (no phosphates) ...... 4 0. Superphosphate, potash, &c ....... 1 A. Ammonia salts.. ................ 2 4. Ditto and ;uperphosphate ........ 3 A. Dittoand potash, &c. (nophosphates: 4 A. Ditto, superphosphate, and p o t d i , Qc. ............................ 1 AA. Siirate of soda.. .............. 2 Ad. Ditto and superphosphate.. ..... 3 A s . Ditto and potash, &c. (no plios- phates) ....................... 4 AA. Ditto, superphosphate, and pot- ash, &c ......................... 1 AAS. Xitrate of s o h and silicate of soda.. ..........................2 AAS. Ditto, ditto, and superphosphate 3 AAS. Ditto, ditto, and potash, &c. (no 4 AM. Ditto, ditto, superphosphate, and potash, Brc. ...................... phosphates) ..................... 1 C. Rape-cake.. .................... 2 0. Ditto and superphospliate ........ 3 C. Ditto and potash, &c. (no plloa- phates) ......................... 4 C. Ditto, superphosphate, and potash, &c.. ............................ $1. Farm-Fard manure for 20 years, un- manure4 last 18 rears ............ T3. Farm-yard manure for 35 years. .... Total phosphoric acid. so~z1tioil of citric acid. lbs. per acre. 2’503 4601 3059 4778 2452 4373 2579 4602 2629 4171 2629 4525 2680 4550 2654 4272 3f31 5408 3590 4794 3332 3669 26s. per nci-e. 499 i.;r70 253 9.360 652 q07.3 205 4264 170 909 20; 620 I 1s 0 4201 233 621 I 4 p 9.303 505 6330 5 12 932 Those closely interested [in the subject will scarcely need any guidance for the study of these tables.Almdst every one of the analytical figures possesses an interest read in conjunction with the history and yield of its own plot, and correlated mich the data, historical and analytical, of other plots. Although, horn ever, the reader having sufficient interest and snffi-PLANT FOOD IN SOILS. 147 cient patience to thus stndy the fignres in detail will need no com- mentative guidance, some of the results may be summarised for the sake of those who care only to learn the main conclusions, if any, to which the investigation tends. In the next table are four averages deduced each from the analyses of four plots, viz.: l(0, A, AA, U S ) , 3(0, A, 88, US), 4(0, A, AA, U S ) , and i(0, A, AA, AAS). YOL. TIXV. MAverage. ~ ~~ Total pliosplioric acid per cent.. ............ acid solution per ccnt. .................. Phosphoric acid soluble in 1 per cent. citric Total 1)hosphoric acid cstimntrd in Ibs. prr acrc.. ................................ Pliosplioric iwid soluble in 1 per cent. citric acid solution estimated in lbs. p w acrc . , . . Rnrleg yield, average per annuni for 38 year3 aw ditto, ditto ........................ Barlcy yield in 1890.. .................... Straw yicltl in 1890 ...................... l(0, A, Ah, AAS), 4 plots. 3 nianured with nitro. gCn. N O MINEltALS. 0 -0063 2566 lbs. 160 lbs. 28a buslicls 10t cwts. 242 buslicls 1zg cwts.3(0, A, AA, AAS), 3 manured with nitro- gcn ; all with potash, soda, und magnesin. 4 plots. - NO PIIOSPIIATES. 0'109 0 *0094 2730 lbs. 237 lbs. 30b bushels 17% cwts. 24# bushels 13; cwts. 4(0, A, AA, AAS), 4 plot,s. 3 manured with nitro- gen ; all with potash, soda, and magnesia, tind WITH riros- PHATES. ---- 0 -180 0.0498 4544 lbs. 1299 lhs. 39# bushels 23$ cwts. 388 busliels 194 cwts. 2(0, A, AA,AAS), 3 manured with nitro- gcn; all WITH PIIOS- &C. 4 plots. - L'HATES. PITO ]lOttl8h, 0,175 0 * 0428 4124 lbs. 1089 lbs. 38& bushels 21g cwts. 36 busliels 17a cwta.PLANT FOOD IN SOILS. 149 It will be seen that the average percentage of total phosphoric acid in the eight plots that receired no phosphates is 0.106 per cent., while on the phosphatically mauured plots it is 0.178 per cect.There is here, of course, a difference that is actually great, its might be expected; seeing that during the 38 years of experiment, nearly 500 lbs. of phosphoric acid per acre must have been removed from the former plots, without any replacement by manuring, while the latter, though yielding not far short of 800 lbs. of phosphoric mid, had during the same time received nearly 2500 lbs. of phosphoric acid as manure. Roughly speaking, the average dxerence to be ex- pected between the total phosphoric acid left in the phosphated and non-phosphated soils would be somewhere about 0.08 per cent., the actual average difference fonnd by analysis being 0.072. But the difference between 0.105 per cent. and 0.178 per cent., great as it is, would have told us practically nothing apart from the known history and yield of the plots.The difference betweet 2500 lbs. of total phosphoric acid and 4500 lbs. of total phosphoric acid, per acre, appears to be immaterial as a measure of present fer- tility for a crop that only requires 20 lbs. per annum for its actual sns tenance. The average percentage, as we have seen, of total phosphoric acid in the eight plots receiving no phosphates was 0.106; in the eight plots receiving phosphates, 0.178. These numbers are nearly in the ratio 1 : 1.7. Now, however, let us consider the percentages of phosphoric acid soluble in the 1 per cent. solution of citric acid. We find that the average percentage thus found in the eight plots receiving no phosphates was 0.0078 ; in the eight soils that received phosphates i t was 0.0463.Thesepercentages are in the ratio of rLeayly i .- 6. The difference in the percentages of phosphoric acid soluble in dilute citric acid is thus comparatively overwhelming. Further, if we compare the four plots t h a t have received neither phosphoric acid nor potash with those that have received potash, soda. and magnesia salts, without phosphoric acid, we find that, although the total phosphoric acid is practically identical in the two sets of plots, yet the four plots treated with alkaline salts have half as much again of citric acid soluble phosphate as the others-in fact 77 lbs. more per acre. Their fertility is also shown to have been greater, and it appears not impossible that t h i s may hare been in som4 degree due to the solvent chemical action of the soluble alkaliue salts on the phosphates of the soil.Again, there is a considerable difference between the avenge citric acid solnble phosphoric acid of the four completely manured plots and that of those which received phosphates without alkalis. The M Z130 DYER : ATAILABLE " MINERAL " former have 170 lbs. per acre more of such soluble phosphoric acid than have the latter, although their fertility, both on the average of p a r s and in the last yield recorded, is higher. It would almost appear as though the presence of the alkaline salts here again exer- cised a solvent action. This apparent solvent action of potash and soda (2 and magnesia) salts, indicated by the presence of more citric acid soluble phos- phoric acid, is suggested not only by the average results, but by detailed examination of each of the sets of plots from which the fore- going summary was averaged. Unfortunately complete miueral analyses have not been made of the crops grown on all the plots, but on two of them, viz., Plots 2 A and 4 A, complete ash analpses have been made of the whole of the grain and sfraw grown in the course of 40 years.Sir Henry Gilbert has been good enough to supply the author with the results of these analyses, from which the actual gain or loss per acre of mineral con- stituents can be calcnlnted. The crops on Plots 1 A and 1 AA have also been analysed, but only for the last 20 years. On these two plots the total mineral losses per acre during the earlier period of experiment can only be estimated, It is true that as far as potash is concerned, such retrospective esti- maks are of little value, but the phosphoric acid can be computed from already existing figures, with a fair approach to certaiat1, as t h e percentage of phosphoric acid found in grain and in straw is far less variable, and less dependent on soil and manuring than that of pot ash .The following table, based in the case of Plot 1 A on 20 years crop analyses, and in the other plots on a complete series of analyses, is particularly interesting, as affording an actual quantitative illus- trttion of facts which have been already considered, but only in the averages of many plots, and by the aid of rough estimates :-PLANT FOOD IN SOLS 151152 DYER: AVAILABLE “MINEFULL” Sir Henry Gilbert has also been good enough to snpply the author with samples of soil from Plot 2 A, drawn in 1882, and samples from 4 A, drawn in 1868 and in 1882. These include not only samples of the surface soil (top 9 inches), but also samples of the second and third depths of 9 inches each.No samples of the subsoil were drawn in 1889. These samples have been analysed, with the following results (p. 153), the results of the 1889 samples being here repeated for com- parison. The total phosphoric acid in the 1882 sample of Plot 2 A is higher than would have been expected, the difference of 0.003 per cent. between it and the 1889 sample representing an accumulation between 1882 and 1889 of only 76 lbs. per acre, whereas the esti- mated accumulation during that time should amount to 378 lbs.of phosphoric acid per acre. But if the phosphoric acid soluble in 1 per cent. citric acid solution in the two samples be compared, it will be found that there is an increase of 0.01 per cent., yepresenting no less than 253 lbs. per acre. The difference between the total phosphoric acid found in thc 1868 sample and that found in the 1889 sample is 0-035 per cent., or 885 lbs. per acre. The estimated accumulation of phosphoric acid on the plot is about 950 lbs. per acre. Curiously enough the 1886 sample seems, as in the case of Plot 2A, to be unduly rich in total phosphoric acid, for the total phosphoric acid figures indicate an accumulation uf 859 lbs. between 1868 and 1882, and of only 25 lbs. between 1882 and 1889, the esti- mated accumulations being 600 lbs.between 1868 and 1882, and 350 lbs. between 1882 and 1889. The citric acid soluble results, however, are more consistent. Between 1868 and 1882 the citric acid soluble phosphoric acid rises from 0.0231 to 0.0334, a rise of 0*0103 or 260 lbs. per acre; while between 1882 and 1889 it rises from 0.0334 to 0.0500, a rise o€ 0.0166 or 420 lbs. per acre. Between 1868 and 1889 the weak citric acid solution accounts f o r 680 lbe. out of the accumulation of 958 lbs. estimated from analyses of crops and the composition of the manures. The subsoil analjses do not appear to suggest that the cropping or manuring have much affected the subsoil below 9 inches from the surface. From the fact that so little citric acid soluble phosphoric acid is found in the subsoils, eyen on these very abundantly phos- phated soils, it would appear that the superphosphate applied does not, to any appreciable extent, sink or accumulate in the subsoil. The figures for Plot 4 A are even more interesting.But to return to the analyses of the more recent samples. No reference has yet been made to the rape-cake plots, 1, 2, 3, and 4 c.--.--------- 1868. Total plioaphoric acid.. ........................... Phosphoric acid solublc in 1 pcr cent. solution o € citric acid.. ........................................ 1882. PhoRphoric acid solublc in 1 per cent. solution of cit)ric acid.. ........................................ 1889. Phosphoric acid sohble in 1 per cent. solution of citric acid.......................................... Total phoriphoric acid.. ........................... Total phosphoric acid.. ........................... Plot 2 A. Ammonia salts and superphosphate. - 1-9 in. - I 0 -170 0.0328 0 -173 0 *0425 10-18 in. 19-27 in. I I 0.126 0 *0010 Plot 4 A. Ammonia salts, superphosphate, and - potash salts, &c. 1-9 in. 0 *147 0.0231 0 *181 0 -0334 I - 0.182 - 1 0.0500 I 10-18 in. 0 0129 0 * 0023 0.125 0 *0015 - - 19-27 in. 0 -129 0 -0014 0 -122 0 *001.6 - -154 DYER : AVAILABLE “ MINERAL” These, as might be expected from their treatment, are far richer in phosphoric acid than the corresponding plots of the other series, though the comparative richness is brought ont far more strikingly in the citric acid soluble proportions than in the total percentages.In the case of the two farm-yard manure plots we have respectively 0.134 per cent. and 0.176 per cent. of total phosphoric acid. Not the most experienced chemist could venture, with no information but these two numbers to guide him, to form any reliable opinion of the comparative phosphoric acid fertility of the two soils; and yet one, after 20 years of dunging, has been left for nearly 20 years unmanured, while the other has been dunged on year by year until to-day. When, however, me compare the citric acid soluble phosphoric acid, we find that the one plot is rather more than twice as rich as the other. Its present fertility, too (though this of course depends also on other constituents than phosphoric acid), is rather more than twice as great as that of the other.SUGGESTED CONCLUSION AS TO PHOSPHORIC ACID. From a careful consideration of the whole of the results, it would perhaps not be unreasonable to suggest that, when a soil isfound to c o d a i n as litile as abovt 0.04 per cent. of phosphoric acid soluble i~ a 1 per cent. solution of citric acid, it would be just$able to assume that i t stands in immediate need of phosphatic manure. POTASH RESULTS. The next tables show the results of the potash determinations, ex- pressed as percentages, as in the case of the phosphoric acid results, and also in the form of estimated lbs. per acre corresponding to those percentages. Like the phosphoric acid results, they are interesting to examine, the differences between the potash found in the plots manured with potash and in those not so manured being, however, even more striking if we regard the proportions soluble in citric acid.PLANT FOOD IN SOILS.155 Potash Determinations i n Samples of Barley Soils from Hoos Field, Rotharnsted. ~~ ~~ ~ I 1 I dry state. Percentages in fine soil calculated on Manure applied every year since 1852 (for quantities see pages 143 and 144). Total potash yielded ~ by acid and subse- quent lime "fusion' of the undissolved I siliceous matter. I 1 0. No manure.. ................. 2 0. Superphosphate.. ............. 3 0. Potash, &c. (no phosphates). .... 4 0. Superphosphates, potash, &c. ... 1 A. Ammonia Ealts.. .............. 2 A. ,, ,, and superphos- phate ........ 3 A. ,, ,, and potash, &c. (no phosphate) 4 a. ,, ,, superphosphate, and potash, &c.1 AA. Sitrate of soda .............. 2A-A. 9 % ,, and superphos- phate ........ 3A.L 9, ,, and potash, &c. (no phosphates) 4 LA.. ,, ,, superphosphate and potash, &c. 1 U S . Kitrate of soda and silicate of , I 2 u s . ,, ,, andeuperphos- I 3 Ads. ,, ), and phate.* potash, ... &c. *I (no phosphates) ' 4 ;LA§. ,, ), superphosphate: and potash, &. 1 C. Rape-cake ..................... 3 C. )) and potash, &c. (no phosplrstes) ....... 4 C. ,, superphosphate and 1 7'. Farm-yard manure for 20 years I (unmanurea for 18 jears) ...... ' ....... soda. 2 C . .. and superphosphate.. .. potash, Be. ........ 7". Farm-yard manure for 38 years.. . I 1 -448 1 *500 1 *695 1'718 1 -418 1 '439 1 '368 1 *713 1 -293 1 -353 1.396 1 *390 1 -423 1.531 1'574 1 *572 1 *384 1 -363 1 *SO4 1 -605 1 -603 1 -601 Potash dissolved by HC1 only.-~ 0 -153 0.204 0 *318 0 *300 0.267 0 -248 0 *257 0 '326 0'136 0 '142 0 -239 0 '210 0 -193 0 -18s 0 '230 0 *%O 0 .l70 0 '194 0 *219 0 '238 0 -159 0 -167 Pofash !issolved by 1 jvw cent. rolution of :ifric acid. -- 0.00.36 0 '0065 0.0366 0 .0&0 0 - 0020 0 '0023 0 ,0407 0 *0298 0 .0050 0'0036' 0 * 0350 0 03a5 0 .OO@ 0 * 0033 O'OqS/, 0 '0870 0 * 0079 0.035.1 0 * 0304 0 -0635 0.0321 0 .0079156 DYER : AVAILABLE ‘‘ MINERAL ” Potash. Percentage results calculated into lbs. per acre, on the assumption that the samples drawn fairly represented the soil to a depth of 9 inches. Mknure applied every year since 1852 (for quantities see pages 143 and 144). ----- 1 0.No manure.. ........ 2 0. Superphosphate.. .... 3 0. Potash, &. (no phos- phates) .............. 4 0. Superphosphate, pot- ash, &... ............ 1 A. Ammonia salts.. ..... 2 A. Ditto and superphos- phate. ............... 3 A. Ditto and potash, &c. (no phosphates). ...... 4 A. Ditto, superphosphate and potash, &c. ....... 1 AA. Nitrate of soda.. ... 2 AA. Ditto and superphos- phate ................ 3 A.A. Ditto and potash, &c. 4 M. Ditto, superphos- phate, and potash, &c. . 1 AAS. Nitrateof soda and silicate of soda.. ...... 2 AAS. Ditto, ditto, and superphosphat e ....... 3 AAS. Ditto, ditto, andpot- ash, &c. (no phosphates) 4 AAS. Ditto, ditto, super- phosphate and potash, &c. .................. (no phosphates). ...... 1 C.Rape-cake ........... 2 C. Ditto and superphos- pliate.. .............. 3 C. Ditto and potash, &c. 4 C. Dittu, superphosphate, (no phosphates). ...... and potash, 8;c. ....... 71. Farm-yard manure for 20 years, unmanured, 72. Farm-yard manure for last 18 years.. ........ 38 years ............. Total potash. lbs. per awe. 36,604 37,918 42,848 43,429 35,845 36,376 39,637 43,301 32,686 34,203 35,290 35,138 35,973 38,701 39,789 39,738 32,683 32,187 35,517 37,902 39,864 33,374 Potash dis- solved by HCI. 4626 5157 8039 7584 6750 6269 6497 8242 34338 3589 6041 5309 48’79 4752 5814 6320 4014 4581 5172 5620 3954 3481 Potash dissolved .3y a l p e r cent.soZution of citric acid. ~~ 94 165 925 859 50 57 4029 753 126 96 884 771 4 06 89 4447 6S3 46 7 4s‘7 829 7 I& 336 669 For brevity’s sake, 1G of the plots, arranged in comparable groups, have been, as in t h e case of the phosphoric acid results, averaged.Average.l(0, A, AA, AAS), 4 plots. - 3 i,lrmured with nitro. gen. No minerals. --------I- Total potash per cent. .................... Potash soluble in HC1 per cent.. ........... Potash soluble in 1 per cent. citric acid solo- tion per oent.. ......................... Total potash estimated in lbs. per acre ...... Potash soluble in HC1 estimated in ibs. per acre.. ................................ .lotash soluble in 1 per cent. citric acid soln- tion estimated in lbs. per acre,, .......... Average barley yield 38 years ............. ,, straw ,, ,, .............. Barloy yield in 1890 ..................... 2(op A' 4 plots.- manured with nitro- gcn ; with l)llOs. phates. No POTASH. St.l'ilW , , ..................... 1.396 per cent. 0.195 ,, 0.003'7 ,, 35,277 lbs. ------_I_---- 1 *456 per cent. 1 558 per cant,. 0.196 ,, 0.261 ,, 0'0040 ,, 0.0394 ,, 36,800 lh. 39,391 1bP. 3(0, A, AA, AAS), 3 mannrcd with nitro- gcn; all with P O T A ~ ~ L , 8odtl, and magnesia. No pliospliates. 4 plots. - 4,924 lbs. 4,942 Ihs. 93 lbs. 28a busbeh 164 cwts. 243 bushels 12a cwt,s. 6,598 Ibs. 102 lbs. 996 lbs. 38H bushels 30h bushel8 213 cwts. 17% cwts. 243 bushels 30 bnsliels 1 7 i cwts. 13A cwts. 4(0, A, AA, AAH), 4 plots. 3 iuanured with nitro- gen; u11 with phos- phates, POTABH, soda, m c l magnesia. - ~~ 1 -598 per cent. 0.272 ,, 0'0303 ,, 40,401 lbs. 6,864 lbs. 767 lbs. 39% bushels 23% cwts. 381 bushels 19a CWt.8.158 DYER : AVAILABLE ‘‘ MINERAL ” The total potash determinations, as might be expected, are devoid of comparative significance, apart from a priori knowledge.The total potash now existing in the eight non-potash plots averages 1.426 per cent. ; in the potash plots 1.578 per cent. These numbers are in the ratio 1 : 1.07. The hydrochloric acid soluble potash averages on the eight non-potash plots 0.195, and on the eight potash plots 0.266. These numbers are in the ratio 1 : 1.36, the difference again being of little or no significance, apart from a priori know- ledge, as any index of comparative practical poverty or richness in potash. Now, however, let us turn to the citric acid soluble percentages. The eight non-potash plots average 0.0038 per cent., while the eight potash plots average 0.0348.These numbers are in the ratio 1 : 9, the difference being, as in the case of the phosphoric acid, compra- tively overwhelming. The difference between the citric acid soluble potash in the four potash-manured plots that have had no phosphates and that in the four potash-manured plots that have been also manured with phos- phates is striking and interesting. The crops on the former, being gradnally starved €or wmt of phosphoric acid, have dwindled down t o 24g bushels of barley and 133 cwts. of straw. The plots fed with phosphoric acid in addition to potash (and nitrogen) have so kept up, that the crops in 1890 still averaged 389 bushels of grain and 192 cwts. of straw. Conseqnantly, these latter plots have removed more potash in the course of 39 gears than the former.This quantity of potash, being unremoved by crops, will have accumulated in the soil. On the four plots in which i t has so accumulated the average quactity of citric acid soluble potash is 996 lbs. per acre ; on the other four plots it is 767 lbs. Thus there is a difference of 229 lbs. of citric acid soluble potash. On the rape-cake plots, which are not included in the average results just considered, me have, as would naturally be expected, a much larger quantity of citric acid soluble potash on the non-potash plots than on the corresponding plots in the various series receiving no rape-cake, rape-cake itself containing an appreciable quantity of potash. Here again, too, the plot that has received potash without superphosphate is richer in citric acid soluble potash than that which has had superphosphate in addition to rape-cake and potash, to the extent of 89 Ibs.per acre. The latter plot has yielded larger crops, which will have removed from the soil much mora potash than those produced on the latter. Much of the difference appears as accumulated citric acid soluble potash. In the two farm-yard manure plots, one of which was dunged for nearly 20 years and nnmanured for nearly 20 years, while the otherPLANT FOOD IN SOILS. 159 has been dunged every year throughout the whole time, the per- centages of potash are respectively 1.603 and 1.601. Duplicate samples of the same soil might well differ to such an extent as that, or even two analyses of the same soil as ordinarily carried out.The respective hydrochloric acid soluble percentages 0.159 and 0.167, obtained by the ordinary mode of analysis, are also to be regarded as virtually identical. Certainly they convey no indication of difference in fertility or condition. But when me examine again the citric acid soluble percentages, me find respectively 0.0135 per cent. and 0.0321 per cent. The continuously dunged plot is more than twice as rich as the other, closely corresponding to the differences indicated also in the citric acid soluble phosphoric acid. The comparative present fertility of the two plots, as has been already pointed oat, endorses the indications thus furnished, for the one yields now more thaiz twice as rnuchproduce as the other. A further examination of the figures conveys more interesting information.I f me turn back to the table indicating the treatment and yield of each plot (pp. 143 and 144), and look down each group of plots, we see that in the 0 group, where no nitrogen is applied, Plots 1 and 2 yield sniall crops. There is no strain on the natural potash resources of the soil ; and we find that these two plots are about twice and three times as rich in citric acid soluble potash as are the corresponding plots of the next (ammonia salts) group. The yield of grain, but more particularly of straw, on the Plots 2 8 and 4 8 shows that the former is now snffering from potash starvation. On analysis its soil is shown to contain oizly 0.0023 per cent. of citric ncid soluble potash. On comparing the yields of grain and straw from Plots 2 U and 4AA, correspondingly manured, except that the nitrate of soda is used as a source of nitrogen instead of ammonia salts, we do not find anything like the same appreciable indications of potash starvation in the plot that has had no potash.On turning to the analytical results, we see that this plot has 0.0038 per cent. of citric acid soluble potash-nearly twice as much as the corresponding ammonia plot. It Seems not improbable that the constant dressings of nitrate of soda have here acted as a solvent of the natural potash of the soil. The same thing, though not to quite the same extent, is noticeable on comparing the corresponding Plots 2885 and 4AAS, which have been treated with nitrate and silicate of soda, where the citric acid soluble potash is 0.0035, just half as much again as in the correspond- ing ammonia plot.The ash analyses of the Plots 2 8 and 4 8 supplied to the author by- I n this (ammonia salts) group we get valuable information.160 DYER : AVAILABLE '' MINERAL " Sir Henry Gilbert, and already referred to under phosphoric acid, enable the following table t o be drawn up. The results of these calculations are very interesting, showing that of the estimated accumulation of 3027 Ibs. of potash per acre 011 Plot 48, 1973 1bs.-ie., about two-thirds of the total estimated accnmulation-is actually found as potash soluble in hydrochloric acid, while 696 Ibs., or about one-fonrth, exists in a form soluble in weak citric acid solutions. Total potash ........................Potash soluble in HC1.. .............. Potash soluble in 1 per cent. solution of citric acid.. ....................... Total potash, lbs. per acre.. ........... Potash soluble in B C1, lbs. per acre .... Potash solrible in 1 per cent. solution of citric acid.. ....................... Barley yield in 38 years.. ............. straw ..................... straw ........................ Barley ,, (yearly average) 1886 to 1890 Estimated removal of potash in 38 years) crops per acre ..................... Potash added in manure in 38 pears per acre. ............................. Estimated loss or gain of potash per acre Expected difference in potash per acre between the 2 plots.. ............... soluble in hydrochloric acid ......... Difference found by analysis in potash Ditto, ditto citric acid soluble potash ., . Plot 26. Ammonia salts and superphospha te. 1 -4.39 0 '248 0 '0023 36,376 lbs. 6,269 ,, 57 ?, 1625 bushels 912 cwts. 32) bushels 17 cwts. 984. lbs. - -984 lbs. Plot 4 6 . Ammonia, salts, superphosphate, potash, soda, and magnesia. - ---- 1 *713 0 '326 0 -0298 43,301 lbs. 8,242 7? 753 7, lM7 bushels 9so cwtx 362 bushels. 203 cwts. 2057 lbs. 4100 Y ) -i-2043 ,) I 3027 lbs. (=0.120 per cent. on dry soil) 1973 lbs. ( = 0 -078 per cent.) 656 lbs. ( = 0 '0275 per cent.) The surface and subsoil samples of Plots 2 8 and 4A drawn in 1868 and 1882, the phosphoric acid contents of which have been already discussed, have also been examined as regards their potash contents. The figures obtained from the different samples are less consistent than the phosphoric acid figures, but it must be remem- bered that the stock of natural mineral potash in these soils is vast, and evidently unevenly distributed, so that errors of sampling are necessarily greater than when we regard phosphoric acid.PLANT FOOD IN SOILS.161 A 0 0 A 0 0 A 3 0 A 0 0 A 0 0 4 0 0162 DYER : AVAILABLE '' NINERAL " Potash is, no doubt, one of the constituents of soil most likely to be modified and rendered available by the action of winter, weather- ing, frost, rain, &c. The 1889 samples of soil were all drawn in autumn, just after the barley crop bad withdrawn its yearly share of available potash, and before the winter had done its work in pre- paring for next year's crop ; so that it is difficult-more difficult than in the case of phosphoric acid-to draw from the figures any fairly plausible suggestion as t o what percentage limit of citric acid soluble potash should be regarded as marking the non-necessity of special potash applications.There seems good reason to suppose that the use of nitrate of soda (and possibly, t o some extent, of superphosphate) helps to bring about by its solvent action on a part of the main stock in the soil a, yearly supply of available potash. Soda salts generally were found by Voelcker in laboratory experiments t o dissolve or liberate potash from soil. In two sets of field experiments on cabbages, carried out by the author in Sussex and Essex, pot'ash salts, as an addition to phosphatic manure and nitrate of soda, produced an abundant increase ; but the substitution of common salt for potash salts on other plots answered just as well, probably owing to the decomposing action of the salt on the compound silicates of potash existing in the soil.The soil economy of potash is probably more complicated than that of phosphoric acid and needs a good deal of farther investi- gation. Probably this limit lies below 0.005 per cent. '' AVAILBBLE " PHOSPHORIC ACID IN ~ ~ A X K J R E S . After what has been said incidentally in the former part of this paper, it is not necessary to say much more by way of preface to some experiments (confirmatory of the views of Stntzer) now to be recorded. They appear to demonstrate, in more detail t,han the pub- lished figures of Stutzer, the uselessness of ammonium citrate as a means of quantitatively measuring the practical value of commercial phosp%ates, and, on the other hand, to further establish the value for such a purpose of a weak solution of citric acid.I n the following experiments with ammonium citrate, 200 C.G. of the sol 11 tion rendered distiuctly ammoniacal were shaken up with 2 giams of the manure or material to be tested, in a stoppered bottle, acd allowed to stand, with occasional shaking, for three days, at a temperntare of 10" C. to 18" C. The figures indicate the qumtity of phosphoric acid dissolved by the citrate for every 100 parts of phosphoric acid present.PLANT FOOD IN SOILS. Ammonium Citrate Experiments. Strength of ammoniacal ammoniun citrate solution .............. Canadian apatite.. .............Spanish phosphate ............. Aruba phosphate. .............. Belgian phosphate ............. Somme phosphate ............. Another Somme phosphate.. .... South Carolina phosphate ....... Another deposit of the same Cambiidge coprolites ........... Raw Redonda (aluminium) phos- phate. ...................... Calcined ditto, ditto.. .......... Bone meal.. .................. Boiled bone meal.. ............. Stea rued bone flour. ............ Basic slag or cinder ............ Perurian guano (Pabellon de Pica) ,, ), (Punta de Lobos) ,, ,) (Lobos de Afuera: .... 1 .. (Huanillos). ..... Fish guano.. .................. - 10 per cent. 'er cent. -- Df total P&,. 1 '55 1 '73 1-58 0 -86 1 -55 0 -94 3 -71 3 -70 4 -02 54 '92 78 -88 15 -62 8 '44 10 -73 21 '86 4.8 '29 10.61 1.5 '46 19 '43 29 *07 20 per cent.'er cent. of total 0 -76 1 -46 2 -87 1 .12 1-71 1 *33 4 -45 3 -91 4 -25 61 -76 84 -7.1. 25 '42 11 -75 13 *S1 43 -58 49.71 14 -4.1 19 -72 22 '73 37 * i 2 -- P,O,. 30 per cent. 'er cent of total 0 *SY 1 -66 3 -58 2 *06 2 -28 1-96 5 -16 3 -45 3 '72 60 -29 78 -46 36 -66 1.5 -25 16 *83 51 *32 55 -49 19 -55 22 -41 27-35 42 -85 -- P,O,. 40 per cent. per cent of total 1 -28 1.89 2 -52 3 -21 2 *55 3 -37 3.79 4'02 56 *68 73 -86 38 -71 18 '31 18 *38 66 -44 55 -99 16 * 81 24 ' 4 0 29 -25 48 *33 -- P,O,. - - 163 50 per cent. 'er cent. of total 1 -88 2 -45 2 -56 3 -17 2-53 3 -23 3.76 4 -u9 8-27 59 -20 38 -71 18 -55 67-51 58-48 16 -00 23 *43 33 -46 62 -24 -- P,Oj . - -. In the citric acid experiments from 0.5 to 1.0 gram of the manure or material was placed in a bottle with 200 C.C.of the solvent, and allowed to stand, with occasional shaking, for three days. VOL LXP.164 Fish guano ............. I DYER : AVAILABLE '' MTNCRAL " Citric Acid Experiments. 10000 Strength of citric acid in solution. ............. Canadian apatite.. ...... Spanish phosphate ...... Aruba phosphate.. ...... *Belgian phosphate.. .... Somme phosphate ....... Another ditto, ditto.. .... South Carolina, phosphate Another deposit of same.. Cambridge coprolites .... Raw Redonda phosphate. Calcined ditto, ditto.. ... Bone meal ............. Steamed bone flour. ..... Peruvian guano :- Basic slag or cinder.. .... Punta de Lobos.. ... Pabellon de Pica.. .. Lobos de Afuera ... Huanillos .......... 0 -25 per cent.Solvent 400 Material - 1 * ____-- Per cent. of total 12 '59 9 -31 26 -06 1 -54 22 -93 16 -31 29-62 29 -09 30 '18 7 -12 8 -36 80 '80 66 *06 47 -44 PZOj. 98 -14 66 -19 72.31 73 -94 0 -50 per cent. Solvent 200 filaterial - 1 * Per cent. of total 11 -85 7 -34 20 -82 1 '75 23 2 4 16 *69 25 '87 26.82 22 -21 8 -49 10 *74 91 -17 71 *52 57 -79 pZo5- 91 -29 61 -06 74 *42 70 -84 89 -07 1 *O Der cent. Solv& t - 200 ~ -- Material 1 * ~~ Per cent. of total p205. 15 -81 10 '73 29 -99 3 '08 30 *36 30 -51 38 -06 34 -46 33 -31 9 *21 16 *06 100 *oo 89 -66 72 .€MI 97 -50 76.67 87 '23 74 -16 91 -46 The ammonium citrate results appear to assign comparatively no value t o ground mineral calcium phosphates. As a matter of fact we know that finely ground mineral phosphates do afford an available, if not an economical, source of plant food, their value being deter- mined mainly by fineness of grinding and specific hardness.A 1 per cent. solution of citric acid, on the other hand, does assign a value-and a graduated value-to them. Such a solution dissolved only (in round numbers) 11 per cent. of the phosphate in Spanish phosphate, and 16 per cent. of that in Canadian apatite, showing a low degree of availability. Softer phosphates-Sommes, Carolinas, and coprolites-give from 30 to 38 per cent. of their total phosphoric acid, indicating a greater availability, which accords well with practical experience ; while ammonium citrate, on the other hand, comparatively speaking, shows scarcely any solubility. Roue meal shows from 90 to 100 per cent.of its phosphoric acid soluble in 1 per cent, solution of citric acid, while the maximum solubility in bones * I n this csse, no doubt, the neutralisation of the acid by carbonate of lime &minished the solubility.PLANT FOOD IN SOILS. 1G5 to strong ammonium citrate only rose tco 38, and sank as low as 8 when only a 10 per cent. solution of citrate was used. Bone meal is generally looked upon as one of the most available forms of undis- solved phosphatic manure, and the citric acid test gives results which clearly accord with experience. Basic cinder, or slag, which yielded 67 per cent. of its phosphoric acid to a 50 per cent. solution of am- monium citrate, yielded nearly 73 per cent. to a 1 per cent, citric mid solution. Various forms of Peruvian guano gave up from 74 to 97 per cent.of their phosphoric acid to the citric acid solution, but only from 23 to 67 per cent. to the strongest ammonium citrate solution. Fish guano yielded 91g per cent. of its phosphoric acid to 1 per cent. citric acid solution (and 108 per cent. to a larger proportionate bulk of weaker solution), while 62 was the highest proportion yielded to the strongest ammonium citrate. I n all these experiments the materials were finely ground-more finely a good deal than in Stutzer’s experiments-and consequently the solubilities here obtained are higher, though the conclusions are in t8he same direction. P. Wagner (Chem. Zeifung, 1886, 10) objected to the szgges- tions of Tollens and Stutzer, to make a 1 per cent. solution of citric acid the analytical test for “a~aililrble” phosphate, on the ground that it did not attack phosphate oE iron and alumina, whereas am- monium citrate does.This contention is borne out by a comparison of the results just, recorded for Redonda (aluminium) phosphate. This, particularly when calcined, is freely soluble in ammonium citrate solution-to the extent of nearly 85 per cent. of its phosphoric acid-while citric acid only indicates a solubility of 16 per cent. Wagner suggests a combination of the two tests, viz., the use of a solution containing per litre 150 grams of citric acid neutralised with ammonia and 10 grams of citric acid in the free state. This suggestion appears to be objectionable, because the object is to ascertain, if we can, what proportion of the phosphate is readily available as plant food ; and if the use of citric acid has any meaxi- ing at all, it lies in an attempted imitation of the acidity of the root sap of plants. That strongly alkaline ammonium citrate estimates the quan- tity of “ precipitated,” “ reverted,” or “ retrograde ” phosphate may be approximately true, though the experience of the pre- sent author is at variance with that of Lloyd (Chem. SOC. Trans., 1882, 308), that “ r;n amrnoniacal solution of ammonium citrate, uo matter what may be its strength, whilst capable of dissolving pre- cipitated phosphafe of lime, does not act either upon mineral phosphates or upon bone ash, and only to a cery slight extent upon raw bow.” The experiments described, however, in the present N 2166 AVAILABLE “MINERAL ” PLANT FOOD IN S0lT.S. paper were made with a much larger proportion of solvent to mnterial than was used in Lloyd’s experiments and possibly on material more finely ground. What i t is desirable t o determine is readily atmilable phosphate ; whether “ reverted ” or not is of no consequence. Phosphoric acid soluble in water is beyond all doubt much more valuable weight for weight than is any other form, however available, since it is immediately diffusible. Many of the pot experiments made to determine the relative availability of different forms of phosphate are, no doubt, to some extent misleading, for in such experiments the manure is often thoroughly incorporated by hand with the soil. In actual farming the manure is drilled or broadcasted and finally harrowed over, or at once buried by the plough. For its further incorporation with the soil it is dependent upon the action of rain. The property of solubility is, therefore, of great value as a means of diffusion, although in any soil but a pure sand the soluble phosphoric acid must be precipitated or fixed long before most of it functions as actual plant food. ‘‘ Soluble ” phosphate will, therefore, probably continue to corn- mnnd a higher price than undissolved phosphste, however “ assimil- able ” the latter ma - b, for the reason that it goes farther as an im- mediate dressing. But the “ insoluble ” phosphates, as the undis- solved phosphates are usually called in England, must be of variable value according to tbeir fineness and softness, conditions of hydra- tion, &., and to class them all together is clearly unreasonable, how- ever convenient to the analyst or however in accordance with English commercial usage. The use of ammonium citrate (t’he reagent ordinarily employed as a means of diagnosing the condition of “ insoluble ” phosphates), being clearly based upon wrong principles, should, however, be abandoned, and the method of Tollens and Stutzer adopted in its place. CONCLUSION. A 1 per cent. citric acid solntim appears, then, to giye indications fairly bearing out the manurial properties of phosphatic materials as recognised by experience in the field ; i t approximates fairly well to the average strength of the natural solvent (root-sap) used by the plant itself; and, tested by the results it gives on soils of known history and condition, it appears likely to afford a not unreliable msans of gauging, as regards the available ‘‘ mineral ” constituents, the probable fertility of the soil itself. The author has already acknowledged the kindness of Messrs.THE MOLECULAR FORMULE OF LIQJJIDS. 167 Sutton and Sons, of Reading, in providing him with pure cult,iva- tions of the various grasses used for determination of sap acidity, and be would here also acknowledge the valuable help rendered to him in the laboratory work by his former pupil and assistant, Mr. P. H. Perry Coste, B.Sc., and by his present assistants, Mr. James Nimmo, F.I.C., Mr. E. H. Robert,s, and Mr. C. H. Allan Bennett. [Note.-It has been very properly pointed out to the author bhat in the case of soils containing iarge quantities of calcium carbonate, an additional quantity of citric acid, corresponding to the quantity of calcium carbonate, might reasonably be added to the solution cf citric acid.-B. D.]

ISSN:0368-1645    

DOI:10.1039/CT8946500115

出版商:RSC    

年代:1894

数据来源: RSC

摘要:

THE MOLECULAR FORMULE OF LIQJJIDS. 167 XVL-Tize Molecular Formulce of some Liqzcicls, as determined by their Molecular Surface Energy. By EYILY Asrox, B.Sc., and WLLLIAM RAMSAY, Ph.D., F.R.S. IN con tinnation of researches published in these Transactions (18933, 63, lOS9), we proceed to bring under the notice of the Society ex- periments on phenol,* bromine, nitric and snlphnric acids, and phosphorus. It has been our object to select liquids which, from their behaviour or from their analogy with others already investi- gated, promised to show a higher molecular weight than that ex- pressed by their ordinary formulae. It would have been interesting to have included sulphur in our list, but unfortunately the viscosity of liquid sulphur at tempei-atures but little higher than its melting poiut precludes observation by the method of registering ascent in a capillary tu be.It may be well here to recapitulate briefly that the variation of surface energy with temperature is, for non-dissociating liquids, a rectilinear function of the temperature, and that the value of the differential coefficients is, on the average, 2.121. Hence, for a normal or monomolecnlar liquid, the numerical value of z, or, as usually dE Et - ELI t t determined, 7 = 2.121, approximately. E represents mole- cular surface energy, or MU)^, where 7 is surface tension; 31, gaseous molecular weight ; v, specific volume ; and (Mc)~, molecular surface. If the value of M has been wrongly chosen, a coefficient differing from, and less than, 2-121 is obtained. In calculating the true value of M, it has been assumed that the principal cause of variation from * The experiments on phenol were made in conjunction with Dr.J. Shields.168 ASTON AND RAMSAT: THE MOLECCZAR FORMULE this general mean is association, and that the association (z) varies only slowly with temperature. Lek M, be the gaseous molecular weight. and :cN, the actual molecular weight of the liquid uuder es- periment,* then, Kt d { yzs(~otq)a} = 2.121, o r daf dt and &Co + r@,,z.)I. - = 2.121. d P d t It is assumed that xf = H/K,, and r(Mo.c)3. - is neglected as too small to affect the resnlt considerably. I n words, it is assumed that the alteration of association with temperature is negligible. It may be seen, by reference to the memoir on molecular surface energy (Phil. Truns., 184, A, 6551, that acetic acid, and methyl and ethyl alcohoIs do not exhibit a very rapid variation of association with temperature, and we therefore regard our assumption as justified for the limirs of temperature within which we determine tbe coeflicients of variation of molecnlar surface energy.In accordance with what has been said, it is easy to calculate the molecular weight of a liquid by the equation where the symbols have the meanings previously attacbed to them. 1c 1. Phenol.-The sample employed was colourless, and had a con- stant boiling point. We have taken as correct the densities as determined by Kopp (Annulen, 95, 312). It was easy to keep the substance liquid in the lower narrow tube, which held the lower part or stem of the capillary tube (see illnstratioa, Zoc.cit., p. '1094), by occasionally warming it with a Bunsen burner. Phenol (r = 0-01843 cm.). We now proceed to give the resuits of experiments. t. M. __I_- ----- 1-736 1-35 1269 ---- ------- 1.899 1'18 110.9 131 7 3-19 09670 27-90 1 589-9 184-0 2-70 09164 22-39 1 4W.6 I 1 9 Where the expression " actual molecular weight " or " mean molecular weight"OF LIQUIDS AND TEEIR MOLECULAR SURFACE ENERGY. 169 t. lo" 6 46.0 The molecular weight of gaseous phenol is 94 ; stud i t is seen that between 131.7" and 184", the value is rapidly approaching that number. In this it agrees with the alcohols ; and as it has been pre- viously noticed that the higher the molecular weight of acids and alcohols, the less the tendency towards association, it was to be ex- pected that a substance of such high molecular weight as phenol would show small tendency to form molecules of great complexity.Experiment has shown the correctness of this expectation. 2. Bromine.-The difficulty in determining accurately the capillary rise of bromine is that its opacity prevents accurate reading of the lerel of the lower meniscus. An approximate result, however, can be found by reading the upper edge of the meniscus; and if, as seems prabable, the height of this meniscus is not greatly altered by rise of temperature from 10" to 80", the results may be accepted as approximating to truth. Bromine (1' = 0.01046 cm.). h. -- rum. 2.4YO 2.230 -_ 78.1 1 1-972 P* 3 -152 3 031 2.917 I-- dynes 1 ergs 40.27 1 552'08 34-68 1 487-98 -- 29.51 1 426.09 1 M.202 -77 184 59 The densities at these temperatures were calculated from Thorpe's determinations (Trans., 1880, 37, 173). If it be supposed that the capillary rise should have been st milli- metre more in each case, the mean molecular weight is altered only to 201*SY, and 183.15 instead of 160, that of Br,. Hence a constant small error in reading would have little influence on the results. Dr. E. P. Perman showed in 1890 (Proc. Boy. Soc., 48, 49), that even at pressures near saturation, and at 15", the vapour density of bromine is normal ; and Paternb and Nasini, in 1888, found that in aqueous and in acetic acid solutions the molecular weight corresponded with Br2. It would appear that even in the liquid state most of the mole- cules possess that formula ; but some sign of association with fall of temperature appeara to take place.3. Nitric acid.-The sample of acid was prepared from nitric and sxphnric acid ; it was then mixed with phosphoric anhydride, and is used, it is not to be understood that the weight in question is that of a definite molecule, but is a mean result due to the fact that the liquid consists of moleculea of different complexity. For example, a molecule of H20 plus a molecule of H801 would have a mean molecular weight of 45.170 ASTON AND BAMSAY : THE MOLECULAR FORMULIE t. 11"-6 4.6 -2 - distilled over a water bath. A bulb was. filled, and when diluted and titrated, it proved to contain 99.8 per cent. of acid. It had a very pale yellow colour, and fumed strongly on exposure to the air.The tube was filled without warming the acid by exhausting a large globe provided with a stopcock, and attaching the neck of t h e latter with indin-rubber tubing to the drawn-out end of the experi- mental tube containing somewhat more than the requisite amount of acid ; on opening the stopcock, the acid evaporated under reduced pressure, and all air was expelled from the experimental tube. The tube was then sealed. The acid turned much darker on heating to 78", and it appeared to contain some nitric peroxide, but, on cooling, it regained its original pale yellow colour. h. -- 4-613 4,182 -- ATz'tric acid (r = 0.01192 cm.). 78-2 1 3.806 x . x x M. 1.681 105.9 --- 1.864 1 117'4 The densities were determined in a pyknometer at 11.6" and at 46-2". It was found impossible to prevent the acid from boiling at 'i8*2", although its boiling point is given as 86".The densitmy at 78.2" was, therefore, estimated by extrapolation. The results show that nitric acid in the liquid state has a molecular weight greater than that expressed by the formula HN03, and probably cousists of a large proportion of molecules of H,N206, mixed with simple mole- cules. 4. Sulphuric acid.-Acid of composition corresponding to the formula H,SO, is unstable, and dissociates on warming into sulph- uric anhydride and oil of vitriol, approximating to the composition 12H28Oc,H20. It was this acid which was employed for the experi- ments, andanalysis showed it to correspond exactly to the above formula. Its expansion was determined with the pyknometer at the tempera- tures at which its ascent in the capillary tube was measured.In calculating its molecular weight, the formula weight of the substance used has been assumed to be (H,SO,,&H,O) = 99.5 ; but, indeed, this is nncecessary, as the equation on p. 168 permits the determina- tion of molecular weight, M, without any assumption. The acid wets the tube easily ; it was sealed after warming in a Sprengel vacuum, so as to expel all dissolved gas. It was not viscous, but ascended easily,OF LIQUIDS AND TBETR MOLECULhR SURFACE ENERGY. 171 Kfromcurve. --- and gave concordant readings. table. The results are given in the annexed Sdpphuric mid (r = 0,01843 cm.). x x M. I 1 } 0.209 I 32.3 x 99.5 182.5 3.275 184.6 3.244 237 -7 I 3 -170 281-0 3.045 ---- I I 1'7342 51.35 763% ---------- 1'6874 49-40 749.7 I -6341 46 *84 724 -9 1-5912 43.80 690-1 - - ~ - - - - - - - ----- J 0.297 0 -599 1 -072 In mapping these results, it is obvious that the first four ralnes of y(Mo)j lie on a straight line ; the mean value of R read from the line is 0.209, and the molecular weight appears to correspond to (H,SO,),.Of course the water preeent may account to some extent for this extraordinary result, but we are disposed to regard it as true; for what is termed the boiling point of sulphuric acid, lying near 360", is in reality its temperature of dissociation under atmo- spheric pressure ; its true boiling point must lie much higher. Moreover, the existence OE complex snlphates such as NaK5(SO4>, implies a molecular weight corresponding to (H,SO,),.It is also worthy of remark that sulphuric acid possesses no appreciable vapour pressure at the ordinary tempeirtture ; for Johnson fouud (Chem. News, 68,211) that no sign of a sulphate could be detected in caustic soda exposed in a vacuum desiccator, containing sulphuric acid, even after R month. The liquid alloy of sodium and potassium also failed to absorb any vaponr from sulphuric acid after exposure for the same length of time. Now, under similar circumstances, a piece of gold leaf exposed to mercury vapour would have been heavily amal- gamated. It would be worth while making similar experiments to detect the possible volatility of such metals as cadmium and zinc. The small decrease of capillary rise with temperature and the corresponding small decrease of molecular surface energy with tem- perature are highly abnormal, and there appears no escape from the condusion that the molecular weight of liquid sulphuric acid is ex- ceedingly high.19'1 x 99.5 6.7 x 99.5 2 -8 x 99.5 - ------172 ASTON AND RAMSAY: THE MOLECULAR FORMULZ Above 130°, however, the molecular weight shows rapid diminn- tion. This may be due to one or to both of two causes, to the disso- ciation of such compound molecules, or to the partial dissociation of the compound into sulphuric anhydride and water. The method of experiment does not allow of a decision; but it is noticeable that faint fumes begin to be evolved from sulphnric acid at about 130", which would imply the latter conclusion, whether the former conclu- sion is true or not.We incline to ascribe the fall of molecular weight to both causes acting in concert. To calculate the moleculay weight directly from the individual observations is in this case not satisfactory, for the differences in h and consequent,ly in -1 are so small that errors of experiment tell very distinctly. But taking the whole range from 10.2" to 132-5", and calculating the molecular weight by the equation the number 3620 is obtained, which is not far removed from 99.5 X 32. But the other method is better adapted to eliminate error in this instance where several observations have been made. 5. Phosphorus.-This substance gave a great, deal of trouble, chiefly in discovering how to fill the tube. The phosphorus was first dis- tilled in a current of carbon dioxide and received i n a tube similar in form to the experimental tube, but having its narrow portion bent at an acute angle, and sealed to the experimental tube; a constricted portion between the two tubes was plugged tightly with glass wool, yet not so tightly as to prevent a stream of dry carbon dioxide from passing. After a sufficient quantity had distilled over, it was melted, and passed th~ough the glass wool into the experimental tube as a limpid liquid with a very pale yellowish tinge.The phosphorus was then melted away from the constriction, which was sealed. The Spreugel pump having been filled with dry carbon dioxide, the upper portion of the experimental tube was constricted, and secured to the Sprengel pump, and a vacuum was made ; the phosphorus was then melted, when a number of bubbles appeared, and were removed by pumping.It was at first thought advisable to boil the phosphortis in the vacuum, but this was found to produce some red phosphorus, and in later experiments the temperature was not raised so high as to cause ebullition. The tube was then sealed at the constriction, and jacketed with the usual vaponrs. The temperature of boiling carbon bisulphide is to3 little removed from the melting point Gf phosphorus to give gDod results ; hence higher temperatures were chosen. Liquid phosphorus does not satisfactorily wet g'lass, and to this also many of our difficulties are to be ascribed. It was only by sinkingOF LIQUIDS AND THEIR MOLECULAR SURFACE ENERGY. 173 t. - 780.3 132-1 the capillary tube by means of the magnet and raising it quickly that trustworthy readings were obtained ; after standing for a little, the cohesion between the phosphorus and the ~1;dss appeared to diminish. We fettr that it cannot be said that the angle of contact between phosphorus and glass is zero, as in the case of the other liquids in- vestigated ; but i t wonld be sufficient for our purpose if the angle is not materially altered within the limits of temperature employed.However this may be, the experiment yields a probable result, for the molecular weight found corresponds to the formula Po, as shown by the following t,able. h. P. Y. y (31 77) 1. I(. x. x x M. - ----- -- - 3-00 1.714 43-09 748 -2 2-55 1.664 3 5 5 6 629 -6 ------- 2.205 0.94 117.0 The molecular weight was here assumed to be 124, or 31 x 4, and we think that the difference is caused by experimental error. At all events, it may be said that there is no sign of association. In conclusion, it may be well to mention here that inasmuch as nitroet,hane was found to associate, an experiment was made with chloropicrin, C(NO,)Cl,, which gave normal results. The liquid is monomolecular. Our results shorn that phenol, like t,he alcohols? forms complex molecules in the liquid statre, which dissociate on rise of temperature ; that bromine, too, is somewhat associated, and aiso gives simpler molecular groupings of Br2 on rise of temperature ; that nitric acid consists largely of molecular groups of the magnitude (HNO,), ; that sulphuric acid is a very complex substance at ordinary tempera- ture, of probably not lower complexity tbati (H,SO,),, but tbat above 132" a rapid diminution of molecular weight is noticeable ; and that phosphorus in the liquid state has a molecular weight equal to that of its gas, P4. University College, London.

ISSN:0368-1645    

DOI:10.1039/CT8946500167

出版商:RSC    

年代:1894

数据来源: RSC

摘要:

174 XVII.-Cositi.ibutions to o w Knowledye of the Aconite Alkaloids. VIIT. On Picraconitine. By WYNDRAM R. DUNSTAN, M.A., F.R.S., and E. F. HARRISOX, Demonstmtor in the Research Laboratory of the Pharmaceutical Society. PICRACONITINE is the name given by Wright (J. Chem. SOL, 1877, i, 143 ; Bee also Wright and Luff, Trans., 1878, 332 and 336) to an amorph- ous alkaloyd obtained in 1874 by T. B. Groves, from the roots of Aconiturn Napellus. The composition and properties of the substance were described by Wright (Zoc. cit.) in 1877. The free base is stated to be a varnish, which could not be crystallised, although, when dissolved in acids, crystalline salts were produced. The compound did not give rise to the tingling of the tongue and lips which is so characteristic of aconitine, but both the base and its salts were intensely bitter ; hence the name propo.sed for the alkalojid.It waa found not to be toxic in small doses. The formula C,,H,,NO,,,, deduced from analyses of the base and its salts, differs from that of aconitine in containing 2 atoms less of carbon and oxygen and 2 atoms more of hydrogen ; Wright’s formula for aconitine being C,H,NO,, The hydrochtoride of the base was represented by the formula C3,HoN0,0,HCl,1QHz0, and the yellow, sparingly soluble, amorphous u,urichZoride by the formula C,HoNOlo,HC1,AuCl3. The actual analytical figures agreed fairly well with those calculated from these formula When heated with strong aqueous alkalis, picraconitine was hydrolysed, yielding bencoic acid and a base closely resembling aconine ; the latter was named picraconine, the hydrolysis being represented by the equation C,,H,NO, + H,O = C7H6O2 + Cf4Ho,N0, Neither picraconine nor asly of its salts could be crystallised.Whilst the base resembled aconine in its chief properties, it w m observed to be rather more soluble in ether and to differ from it in composition ; the analyses of the base and of its mercuro-iodide agreeing fairly with the formula CzaHuN09, which contains H, more and CzO, less than the formula f c)r aconine. Some mystery has always surrounded this alkalo’id on account of iis having only once been obtained from the roots of A . Napellus. Groves, who had many times previously extracted the total alkaloyds from t.he roots of this plant, had not before obtained any such sub- stance, and, moreover, neither he nor Wright were able to obtain i t from undoubiedly genuine specimens of A.Nupellus. Later observers, also, have failed to find picraconitine, and the suggestion has beenDUXSTAN AND HARRISON ON PICRACONITILW. 175 made that, in all probability, it was derived from the mots of some other species of aconite than A. Napellus, which were present in the collection on which Groves originally operated. After reading Wright’s description of the properties of picr- aconit<ine, it occurred to us as probable that it miqht be isaconitine, which we have shown to occur in A. Napellus, and to be producible from aconitine, contaminated with a small proportion of other base. A11 the properties attributed by Wright to this alkalo‘id are con- sistent with this view as to its nature, the presence of impurity giving rise to the analytical numbers which this observer recorded, and also t o the slight difference he noticed between sconine and picraconine.Moreover, the great difficulty that there is in completely purifying isaconitine rezdered it extremely likely thaz i f this alkaloyd had been isolated it would not have received sufficient purification to make clear its real nature and composition. It was obviously not at all easy for us to verify this hypothesis as to the cornposition of picraconitine. Dr. Alder Wright, who has rendered considerable assistance to our enquiry by putting authentic specimens of his products at our disposal, was unable to help us on this occasion, as all the picraconitine he possessed was used up in making the analyseq referred to above.We next applied to Mr. T. B. Groves, who very kindly searched his collection of aconite pro- ducts, aud most fortnnately discovered among them small quantities of picraconitine salts, which he was good enough to send us for the purposes of examination. One specimen, weighing less than half a gram, was labelled “Picraconitine muriate” ; the other, of which there was a similar quantity, was labelled “ Picraconitine nitrate.” The picraconitine hydrochloride was a whitish powder, having a very bitter taste, and produciug scarcely any tiuglirig sensation on the tongue and lips ; it melted somewhat indefinitely near 2 1 2 O , arid on decom- posing a portion of it with ummonia a base was obtained soluble in ether, and presenting the appearance of a varnish which refused to crystallise.The salt was recrystallised from hot water in the manner described in a previous paper in connection with the purification of isaconitine hydrochloride ; as, however, the crystals still melted at rather a lower temperature than the pure isaconitine salt, they were again recrystallised from a mixture of alcohol and ether, from which they separated quite characteristically. The crystals now melted at the same temperature as isaconitine hydrochloride, with which they corresponded in every respect. The specimen of picraconitine sritrate was sligbtly yellowish, and melted near 170”. On recrystallisation from water, it separated as a white, crystalline powder, which melted indefinitely, with decomposition, near 212O.The addition of ammonia to an aqueous176 DUNSTAN AND CARR: THE EFFECT OF HEAT solution of the mlt precipitated the base in white flocks, which dissolved i n ether and remained as an uncrjstallisable varnish on evaporating the solvent. The base thus obtained was converted into hydrochloride, and, like issconitine hydrochloride, the salt separated from a hot solution during evaporation ; i n order to com- pletely purify the salt, i t was several times recrptallised from a mix- ture of alcohol and ether, when it appeared as rosettes of silky needles, which finally melted a t the same temperature as pure isacon- itine hydrochloride, 217". There was thus little room for doubt that we were dealing with a salt of the alkaloid we have previously described under the name of isaconitine. As a small quantity of material remained, it was decided, in order to make the proof conclusive, to prepare the very charac- teristic aurichlor-derivative of isaconitine, C,Ho,( AuC12)N0,,. The addition of a solution of auric chloride to a solution of the hydro- chloride produced a pale jellow, amorphous precipitate, which was dissolved in warm absolute alcohol and the solution allowed to stand. No crjstalline auiichloride separated. Ether was now added to the solution and the mixture kept for a day in a stoppered bottle ; small, colourless crystals gradually separated, and these melted a t the same temperature as pure aurichlorisaconitine, namely, a t 204". These observations prove t,hat " picraconitine " is only impure is- aconitine, and t h a t , therefore, this name can no longer be employed to designate an alkalo'idal constituent of Aconitum Xapellus. Research Laboratory, Pharmaceutical Society, London.

ISSN:0368-1645    

DOI:10.1039/CT8946500174

出版商:RSC    

年代:1894

数据来源: RSC

摘要:

176 DUNSTAN AND CARR: THE EFFECT OF HEAT XV I~I.-Co?itrib~itio7.zs to o w Kyzozoledge o f the Acotzite Part IX. Tlze Eflect qf Heat on ACOW Formation o j Alkaloids. itine und some o f its Deiiuatives. Pyraco&tin e. By WYNDHAM R. DUNSTAX, M.A., F.R.S., and FRANCIS H. CARR, Assis- tant Demonstrator in the Research Laboratory of the Phsrma- ceut ical Society. IN Part I of these contributions, the observation was recorded that when aconitine is heated at its melting point, 188-189", a volatile compound escapes, the alkaldid losing about 10 per cent. in weight. At first i t was conjectured that this might be due to the loss of benzoic anhydride and the formation of aconine. Further investiga- tion has shown, however, that the decoqposition is of a whollyON BCONITINE AND SOME OF ITS DERIVATIVES.177 different kind, and that a new alkaloid is prodnced which, having r e p r d to its mode of origin, we propose to namepyraconitine. By heating in a para6n bath a weighed quantity of pure aconitine at its melting point (189-190"), the alkaloid being contained in a small bulb-tube, having a smell U-tube sealed on so as to act ZUJ a condenser, it was ascertained that acetic acid distilled over. By con- ducting the last stage of this operation in a slow current of air, it was proved, by titrating the distillate with standard alkali, that 9.28 per cent. of the acid is given off from the alkaloid ; the loss, deter- mined by weighing the residue, is represented by almost the same percentage, namely 9.96 ; weight of aconitine, 1.5079 ; weight of residue, 1.3577 = 9.96 per cent.The loss of 1 molecular proportion of acetic acid by 1 molecular proportion of aconitine corresponds with 9.27 per cent., a number which is in close agreement with the experi- mental results. A minute quantity of benzoic acid also passes over, b u t the amount is almost inappreciable if the temperature is not allowed to rise during the experiment. The identity of the acetic acid was conclusively proved by conrerting it into the silver salt and submitting this to analysis. Silver salt, 0.1722 gram ; silver, 0.1112 gram = 64.58 per cent. Calculated for silver acetate, 64-66 per cent. The production of acetic acid from aconitine was not to be inferred from previous knowledge of the properties of this alkdoid. Wright has stated that benzoiu acid is the only acid formed when aconitine is hydrolped.This benzoic acid which so readily separates on hydrolysis is not detached from acouitine when it is heated at its melting point. There is no doubt that aconitine contains one benzoyl group, and that the acetic acid is derived from an acetyl group. Further investigation has shown that aconitine must be regarded as acetyl- benzoylaconine, and our so-called " isaconitine " as benzoylaconine ; for we shall prove in our next paper that one molecular propor- tion of acetic acid separates when the one alkalo'id changes into t h e other, and also when aconitine is hydrolysed. Composition and Properties of Pyraconitine. The dark-coloured alkalo'idal residue remaining after the distillation was dissolved in dilute sulphnric acid and the solution fiitered; dilute ammonia was then added in slight excess, and the liquid extracted several times with ether.The alkaloid was removed from the ethereal solution by shaking it with very dilute hgdrobromic acid, and the acid liquid precipitated in fractions by the addition of dilute ammonia in order to remove the brown colouring matter, which1 i 8 DUNSTAN AND CARR: THE EFFECT OF HEAT accumulated in the first and second precipitates ; excess of ammonia was then added to the colonrless solution, and the alkaloid extracted with ether. When the dry ethereal solution was spontaneously evaporated, it left a vamish-like residue of base which refused to crgstallise ; i t was therefore dissolved in dilute bydrobromic acid, and the neutral liquid slowly evaporated in a desiccator.N o crystal- line salt separated. The dry residue was next dissolved in absolute alcohol, and the clear solution mixed with anhpdrous ether until it became slightly turbid. The small, white crystals of the hydrobro- mide which separated from t'his mixture melted at 280", and still melted at the same temperatnre after recrystallisation. On adding soda in slight excess to the solution of the pure salt, and extracting with ether, white slender needles (m. p. 167.5" corr.) of the base were ohtained by the slow evaporation of the ethereal solution. The crystals are sparingly soluble in water, but readily in alcnhol. chloroform, and ether. The aqueous solution is slightly bitter, but with a, somewhat burning after-taste.The solution in anhydrous alcohol is without action on polarised light ; althouzh some specimens examined, probably not quite pure, exerted a feeble dextrorotation, and others a feeble hvorotation. The base readily dissolves in acidq. forming Falts which are difficult to obtain well ~rpst~allised from an aqueous solution. but separate in well-defined crystals from solution in a mixture of anby- dlrous alcohol and ether. Pyraconitine hydrobromide C,lH,,NO,,,HBr separates in anhydrnns crystals from a mixture of alcohol and ether, and melts at 280" (corr.). It is soluble in water and in alcohol. 0.2289 of salt gave 0.0624 AgBr. C31H41NOloHBr requires Br = 11.97 per cent. On some occasions crystals of this salt., possibly hydra edi have been obtained, melting at 204".When burned in a stream of oxygen, the salt furnished the following data. 0.1452 gave 0.2975 CO, and 0.08399 H,O. C = 55.87 ; H = 6.38. 0.1314 ,, 0.268-5 ,, ,, 0.0731 ,, C = 55.72; H = 6.18. Calculated for C31H11NOlo,HBr. Br = 11%. C = 55.69 ; H = 6.28 per cent. 7 9 ,, C3,H,,N0,,,,HBr. C = 55-85 ; H = 6.0 ,, Of the two probable formuls for aconitine (Part I), namely, C,H,,NOI2 and C33H45N012, these analyses agree better, perhaps, with t h e formula for pyraconitine derived from C31H,,NOlo. However, until more decisive evidence has been obtained on this point we shall retain tbe formula C,,H,lNOlo. The equation representing the production of pyraconitine from wonithe is thereforeON A C O " E AND SOME OF ITS DERIVATIVES. 179 c,@ao,* = C*H40* + Cs,H,,NO,o Aconitine.Acetic acid. Pyraconitine. Pyraconitine hydrobromide is hvorotatory. A determination of this specific rotation in aqueous solution gave the following result. a[15"] = -2"; Z = 2 dcm.; c = 2.1364, whence 2 x 100 2 x 2.1368 ______ = [a]= = --46*fi0. Pyraconitine hydrochEm*'de, C3,H,,N Ol,oH C 1, crystallises in rosettes from a mixture of alcohol and ether. Pyraconitine hydriodide, Cal H4,NOI0,HI, also crystallises from a mixture of alcohol and ether in colourless rosettes, which are apt to become yellowish on standing. They melt sharply at 220.5" (mrr.), and can be crystallised (in prisms) from aqueous solution more readily than the hydrochloride o r hydrobromide. The salts of pyraconitine are somewhat bitter, and are apparently not toxic in small doses.The exact physiological action of this alkaloid is being investigated by Professor Cash, F.R.S. Pyraconitine aurichh-ide, C31H~lNOlo,HAuC14, is thrown down as a pale yellow precipitate, when auric chloride is added to a solution of the hydrochloride ; all attempts to crystallise this salt have hitherto failed. Experiments have been made with the view of preparing an aurichlor-derivative from it similar to that producible in the case of isaconitine, but the pyraconitine compound exhibits no tendency to form such a derivative. It melts at 248.75" (con.). Hydrolysis of P yraconitiue. Pyraconitine and its salts readily undergo hydrolysis when heafed with water and acids, or when allowed to remain in contact with fixed alkalis. When soda is added to an aqueous solution of a salt of pyraconitine, the alkaloid is thrown down in white flocks, which gradually dissolve if excess of alkali is added, owing to the rapidity ivith which hydrolysis occurs ; nothing is obtained from the solution on distillation, but ether extracts from it an alkalo'id p y r m & p , which remains as an uncrystallisable varnish when the ether is evaporated. The residual alkaline solution furnishes benzoic a&j when acidified, but no other acid product ; by estimating this, it has been found that the benzoic acid produced amounts to 17.91 per cent.of the pyraconitine. The amount calculated from the equation C,iH,,NO,o + HzO = C,H,Oz + cd&NO, Pjraconitine. Benzoic acid. Pyraconine. is 18.26 per cent. It therefore appears that the benzoyl group of aconitine is present in pyraconitine, which, like aconitine, undergoes VOL.LXV. 0180 DUNSTAN AND CARR: THE EFFECT OF HEAT hydrolysis, losimg its benzoyl group as benzoic acid, and furnishing a new base which it will be convenient, following the analogy of acon- itine, to name ppconine. Compositwn and Properties of Pyraconine. Pyraconine is an amorphous base resembling aconine in its proper- ties. I f is readily soluble in water and also in ether : in the latter respect it differs from aconine. The aqueous solution has a somewhat sweet taste, and is lzvorotatory. A determination of the specific rotation gave the following data. a[l5O] = -2.04"; Z = 2 dcm. ; c = 1.121, whence Pyraconine combines with acids, forming salts which are very soluble in water.They crystallise more or less readily from their aqueous solutions, but are more easiIy obtained in well-defined crystals from a mixture of alcohol and ether. Pyraconine hydrochloride, C2,H,N09,HC1, crystallises in cubes with 1H,O, which is lost at 100". They melt at 154" (corr.), and are very soluble both in alcohol and in water. The aqueous solution is laevo- rotatory. In determining the specific rotation the following results were obtained. a[1S0] = -4.4"; 1 = 2 dcm. ; c = 1.9595, whence loo 4*4 = [z]D = -102.07. 1.9595 x 2 Unlike aconitine, isaconitine, and aconine, the specific rotation of pyraconine and its salts have the same sign, both are laworotatory. This salt was burned for the determination of the carbon and hydrogen, and the water and chlorine were also estimated, with the following results.0.1266 gave 0.2462 CO, and 0.0881 H,O. 0.1097 ,, 0.2154 ,, ,, 0.078 ,, C = 53.55; H = 7.9. 0.2924 lost 0.0087 of water. 0-2837 gave 0-0756 AgC1. C = 53.03 ; H = 7.73. H,O = 2.97. C1 = 6.57. C24H3,NOg,HCl,H20 requires C = 53.59; H = 7-44; H,O = 3-37; Cl = 6.58 per cent. Pyraconine aurichloride, C24H3iX09,HAuC14, is thrown down as a yellow, amorphous precipitate, which quickly aggregates to a sticky mass, when auric chloride is added to a strong solution of pyraconine hydrochloride. I f this salt crystallises at ail, i t does so only with great di5culty. Crystals hare only once been obtained, and thenON ACOhlTINE AND SOJdE OF ITS DERIVATIVES. 181 not in quantity sn9icient to enable us to =certain that they con- sisted of the anrichloride.As a rule, it separates from solvents as an oil. From the observations recorded in the foregoing account, it willbe seen that pyraconitine bears little resemblance to aconitine, but ex- hibits some affinities with isaconitine ; the salts of pyraconitine, how- ever, melt as a rule at lower temperatures, and show greater rotatory- power, than these of isaconitine. Moreover, pyraconitine does not possess the peculiar property of forming an aurichlor-derivative so characteristic of isaconitine. Ppconine, the product of the hydrolysis of pyracouitine, some- what closely resembles aconine, the hydrolytic product of both acon- itine and isaconitine. It differs from aconine, however, in seveml respects, notably in its rotatory power, and that of its salts, as well &s in its solubility in ether.Action of Heat on Aconitine Salts.-We find that the salts of acon- itine, like aconitine itself, lose acetic acid when they are heated at about 190". The acetate of aconitine requires rather a higher temper- ature for its decomposition, acetic acid not being evolved until the salt has been heated t o a temperature of nearly 198". Action of Heat on Isaconitine.-HaTing fully made out the mode of action of heat on aconitine, it became of considerable interest to ascertain how far isaconitine resembled it in this respect. About 1 gram of the pure colourless amorphous base was gradually heated in a paraffin bath. At lls", the alkalo'id fused to a clear liquid, which underwent nothing more than a darkening in colour on raising the temperature to about 200" ; slightly above this decomposition set in, attended with effervescence, and a colourless liquid distilled, consisting chiefly of water; no acetic acid was obtained, in fact, the distillate was not acid to test-paper. A f nrther important constitutional difference between aconitine and isaconitine is thus broiight to light, which is explained by the results recorded in a subsequent paper.Actiow of Heat on Acouine.-As the usual formula of aconine, C26H41N011, differs from that of pyraconine, C24H3,N09, by the elements of acetic acid, C,H,02, and, moreover, as, according t o Wright, only aconine and benzoic acid are obtained by the hydro- lysis of aconitine, aconine when heated should lose acetic acid and furnish pyraconine, for we have shown that aconitine when heated loses acetic acid forming pyraconitine, which, on hydrolysis, loses benzoic acid producing pyraconine.When pure dry aconine is gradu- ally heated in a paraffin bath, it fuses to a clear liquid at 120"; as soon as the temperature has risen to 190", the liquid darkens, and at a somewhat higher temperature begins to effervesce, the vaponrs evolved 0 2182 SCHUNCK AND MBRCHLEWSKI condensing in the receiver to a colonrless liquid. On examination, this was fonnd to be chiefly water; no acetic acid was present, the liquid not being acid to test-paper. By weighing the dark-coloured residneof alkaloid it was ascertained that the aconine had lost nearly 10 per cent. in weight; this residue, when heated at about 2-50’, decomposed completely. The fact that aconine does not give acetic. acid on heating points to the conclusion that this acid must have been split off in its formation from aconitine. We have shown in previous communications that both aconitine and isaconitine undergo hydroly- sis, aconine being the basic product in each case. In a subsequent paper we show that when aconitine passes into isaconitine, it loses acetic acid. Synthetical experiments are being made to test the accuracy of this new view of the constitution of “ aconitine” and “ isaconitine,” which would represent pyraconitine as the anhydro- derivative of the latter alkaloid. Research Laboi*ntory, Phamnaceutical Society, ~ ~ ~ ~ a O g ~ .

ISSN:0368-1645    

DOI:10.1039/CT8946500176

出版商:RSC    

年代:1894

数据来源: RSC

摘要:

182 SCHUNCK AND MBRCHLEWSKI XIX. -ATotes on Mudder Colou&ng Mattew. By EDWARD SCHUSCK, Ph.D., F.R.S., and LEOX MARCHLEWSKI, Ph.D. I. CONSTITUTION O F RVBIADIN. Irr our prexrious communication (Trans., 1893, 63, 969), we gave an account of the properties of this substance, and arrived at the conclusion that it must be considered as a fiornologue of purpnyo- xanthin, which it closely resembles. As formerly mentioned (Eoc. cit., p. 1137), experiments were made with a view t,o determine the constitution of rubiadin, the chief point being to ascertain the position of the methyl group in the molecule. These showed that the methylpurpuroxanthin [CH, : OH : OH = 3’ : 1 : 31 obtained, though very similar to, is not identical with, rubiadin, as its melting point is considerably lower than that of the latter.We are now able to confirm the views formerly expressed as regards the nature of rnbiadin, more especially as to the character of the methyl complex contained in it. The analyses given in our first memoir show that the composition of the substance is undoubtedly that of a dihydroxyanthraqninone containing an additional CH,. Now this CH2 might either be a constituent of a methoxy-group, or it might be a constituent of a methyl group united to a carbon atomON MADDER COLOURINGF MATTERS. 183 of the anthraquinone complex. In order to decide between these two views, we treated rubiadin with hydriodic acid by Zeisel’s process, and found that it did not yield methylic iodide, and could not, there- fore, contain a methoxy-group. This conclusion was confirmed by the following experiment.A small quantity of rubiadin was heated with concentrated sulphuric acid for 15 minutes at 180” ; the solution, when cold, was poured into water, the product extracted with ether, and the ethereal solution evaporated. The residue, when dissolved in benzene, gave crystals of unchanged rubiadin, which melted at about 290”. These experiments proved the absence of any methoxy-group, and it now remained to determine the position of the methyl group in the molecuie. On oxidation, rubiadin yields phthalic acid, so that the methyl group must be contained in the same nucleus as the hydroxyl groups. The experiment was conducted as follows. Rubiadin was boiled with a mixture of glacial acetic acid and chromic acid, and the solution, when cold, was poured into water, and extracted with ether; the ethereal extract, on evaporation, left a greenish mass, which was again extracted with ether, and the filtered solution again evaporated.In this way a small quantity of a colourless substance was obtained, which sublimed easily. The sublimate melted at 128”, and on being cautiously heated with resorcinol, and the product dissolved in caustic soda, gave the characteristic green fluorescence of fluoresceln, proving that it was pht,halic acid. Rubiadin must, therefore, have one of the two following formulae. CO OH CO OH TVhich of these two is the true one could only be decided by a sgnthetical experiment. I n order to obtain a methylpurpuroxanthin having the constitution [OH : CH, : OH = 1 : 2 : 31, we employed benzoic acid and metadihydroxyparatoluic acid.The condensation of these t x o substances in presence of sulphuric acid would take place in accordance with the following equation OH CO OH The dihydroxyparatoluic acid required was prepared by ourselves f ~ o m Kahlbaum’s pamtoluic acid ; its melting point was found to be lis”, as given by Weinreich (Ber., 20, 98.2). The condensation184 YCHUNCK AND MARCHLEWSKI was effected in the manner described in a previous paper by one of us (Zoc. cit., p. 1142), but the heating was continued longer, as i t was found that condensation took place with some difficulty. A mix- ture of 4 grams of dihydroxypaiatolnic acid, 15 grams of benzoic acid, and 200 grams of snlphuric acid was heated for 15 hours at 110-120".The solution was then poured into water and shaken with ether ; the ethereal extract was evaporated, and, the residue having been sus- pended in water, the excess of benzoic acid present was driven off by means of a current of steam. This was~followed by treatment with boiling benzene, and the solution, on cooling, deposited crystalline needles, which, after recrystallisation three times from hot benzene, yielded the following results on analysis. 0.1201 gave 0.3113 CO, and 0.0448 H,O. C15HloOa requires C = 70.86; H = 3.93 per cent. At 270", the product begins to sublime, and melts at about 282", but if heated qnicklyit melts at 290°, which is the melting point of rubiadin itself. The solution in concentrated sulphuric acid is brownish-red, and shows a narrow absorption band in the red, which is a little nearer the less refrangible end of the spectrum than the correspondirg band of a purpuroxanthin solution The substance is easily soluble in alcohol or ether, and is obtained in orange-coloured needles on evaporating the ethereal solution.I t s properties do not therefore differ greatly from those of rubiadin, but seeing that there was some uncertainty as to its precise melting point, an uncertainty due to its beginning to sublime before complete fusion, we determined to find some other means of com- parison between the natural and the artificial products, and for this purpose had recourse to the acetyl derivatives. C = 70.69 ; H = 4.13. It dissolves in alkali, forming a red liquid. Acet y lrubiadiu. I n order to acetylate rubiadin i t was digested with acetic anhydride and it small fragment of stannous chloride.The excess of acetic anhydride was decomposed with alcohol, and after evaporating the ethylic acetate, the residue was crystallised from alcohol. In this way a product was obtained crystallising in lustrous, silky needles, and melting at 225" ; it is not changed by cold alkalis, but is decom- posed on heating with them. Acetyl Compound of the Nethylpurpuroxanthin [OAc : CH, : OAc = 1 : 2 : 31. This compound was prepared in the same way as the preceding. After two crystallisations from alcohol, it melted at 217-218", and differed in appearance from acetylrubiadin, the latter crystallising inON MADDER COLOURING MATTERS. 185 fine needles, whilst the former appears in crystalline p i n s .Hence it follows that the second of the two formulse given above is the one belonging to rnbiadin. This result is of some interest, for as pur- puroxanthincarboxylic acid is held t o be a derivative of a-anthra- quinonecarboxylic acid, and may therefore have the following formula, CO OH it is possible that rubiadin may be the parent substance of this acid, which would therefore be formed from rnbiadin by some process of oxidation going on in the plant. JI. THE MONALKYL ETHERS OF ALIZARIN. Several years ago a monomethylalizarin and a monethylalizarin mere prepared by one of us by heating alizarin with methylic iodide or ethylic iodide and caustic potash, in sealed tubes (Munchester Memoirs, 1873). During this process, strange to say, no dimethylalizarin or di- ethylalizarin is formed, from which it follows that the etherification of one of the hydroxyls of alizarin is not easily effected.This behaviour is not singular. In their memoir on the derivatives of monhydroxy- xanthone (Bey., 1893, 76), Kostanecki and Dreher stat'e that whilst the 2-, 3-, and 4-hydroxyxanthones may be easily methylated, l-hydr- oxyxanthone, in which the OH group adjoins the carbonyl group, remains unchanged when treated with potash and methylic iodide. The same may take place in the case of alizarin, in which the hydr- oxyl group adjoining the carbonyl group is probably the one which is not easily affected. It would appear therefore that the constitution of monomethylalizarin is represented by the following formula.CO OH It melts at 228-229", and may be easily acetylated by Lieber- The acetyl compound crystallises from alcohol in It yielded, on analysis, C = 68-57 ; H = 4.21. mann's method. light yellow needles, and melts at 186-187". the following result's. 0.1187 gave 0.2984 CO, and 0.0450 H20. C,,H,O2(0CH3).O*C2H3O requires C = 68.92 ; H = 4-03 per cent. We endeavoured also to prepare the second possible methylalizarin186 ON MADDER GOLOURING MATTERS. from ruberythric acid. If we assume, with Liebermann and Bergami, that ruberythric acid contains one of the two hydroxyls of alizarin in a free state, and assuming, further, that the union of the alizarin with the sugar takes place through that hydroxyl which may be easily alkylised, it might be possible to obtain a methylrnberythric acid, and, by decomposing the latter, to arrive a t the second methyl- alizarin sought for.Our experiments failed; we were unable to obtain a, methylruberytbric acid by treating the potassium salt with methylic iodide and methylic alcohol. It is, of course, quite possible t h a t , under certain conditions, especially such as occur in t,he rege- table organism, a compound of this kind may be formed, yielding, as a product of decomposition, the monomethylalizarin of Hiimmel and Perkin (Trans., 1893, 63, 1174). BcetyZcfhyInZizarin.-E thylalizarin melts at 188-189", and may be easily acetylised ; the product crystallises from alcohol in yellow needles, and melts a t 141". Analysis yielded the following results. 0.1130 gave 0.2878 C02 and 0.0482 H,O.C = 69.46; H = 4-73. C1PH~OZ(OC2Hj)*0*CzH30 requires C = 69.67 ; H = 4.51 per cent. 111. RCBERYTHRIC A C I D . In connection with previous work on the glucosides (Zoc. cit., p. 1137), which led to the conclusion that the constitution of these compounds is in harmony with Tollens's glucose formula, we examined the action of phenylhydrazine on the glucoside of alizarin. After several ex- periments, we came to the conclusion that the two substances do not react, a fact which points to the absence of aldehyde groups in ruberythric acid. The formula of this glucoside must, therefore, also be in accordance with Tollens's glucose formula, or a similar one; it may, at the same time, be mentioned that, according to Liebermann and Bergami, the sugar is contained in this compound as a biose residue.Ruberythric acid, according to Liebermann and Bergami, yields an octacetyl derivative. We have succeeded in obtaining benzoyl derivatives also, using €or this purpose the method adopted by Baum, Banmann, and Schotten. On shaking up a solution of ruberythric acid in caustic soda (1 part of soda to 8 of water) with benzoyl chloride, a yellow precipitate was formed, which, after being collected, washed first with dilute caustic soda and then with hot qater, was dried and dissolved in benzene. On partial evaporation of the benzene, the benzoyl derivative separated in dotty masses, which appeared amorphous, even under the microscope. Its composition was that of a heptabenzoylruberythric acid, as the following analysis shows.INTERACTION OF BENZYLAMINE, ETC. 187 0.1521 gave 0.3863 C02 and 0.0658 H20. C,5H,02, requires C = 69.65; H = 433 per cent. The compound is insoluble in dilute alkali, but, on boiling, it gradually dissolves, and the solution then contains alizarin, benzoic acid, and sugar, It is decomposed by hot dilute sulphuric acid, yielding the same products. A quantitative determination of these products of decompasition was not thought necessary, as the number of benzoyl groups was sufficiently indicated by analysis. I f a weaker solution of alkali (1 part NaHO to 10 parts H20) be used, a hexa- benzoyl derivative is obtained, as shown by the following analysis. C = 69.29; H = 4.81. 0.1349 gave 0.3396 CO, and 0.0.529 H,O. C6,H,,0zo requires C = 68.60; H = 4.33 per cent. Rubery-thric acid, therefore, behares like glucose as regards the benzoyl derivatives which it yields, different products being obtained according to the concentration of the soda lye emploFed (v. Skranp, Xufiats. Chem., 10, 389, and Baumann, Bey., 19, 3219). C = 68.65 ; H = 4.36

ISSN:0368-1645    

DOI:10.1039/CT8946500182

出版商:RSC    

年代:1894

数据来源: RSC

摘要:

INTERACTION OF BENZYLAMINE, ETC. 187 XX,-l% temct ion of Benxylunzine aid Etlz ylic Chlor- ucetate. By ARTHUR T. ~IASON, Ph.D., F.I.C., and GOODLBTTE R. ~ I S D E R , Ph.l)., F.I.C. AJIIDOAC ETIC acid (glycocoll) and methylamidoacetic acid (sarkosine), as well as their ethylic salts, easily lose the elements of water, t'orming so-called anhydrides, which, as Curtins and Schulz have shown (Ber., 23, 3041), are derivatives of piperazine, formed by the union of 2 mols. of the compounds with elimination of 2 mols. o€ wafer or of alcohol, the substances being ay-diacipiperazine and dimethyl-ay-diacipiper- azine respectively. Rebnffat (Gazzetta, 1'7: 231 ; Bey., 24 Ref. 136) has shown that the chief product of the interaction of aniline and chloractic acid is phenylglycocoll, C,H,*NH*CH,*COOH ; by heating the latter at 140-150", it is converted into a compound of the formula (Hausciorfer, Ber., 22, 17971, which i s identical with P.J. Meyer's phenylglycine anhydride (Ber., 10,196?), and with Abenius' diphenyl-188 MASON AND WINDER: INTERACTION OF ketopiperazine (Ber., 21,1665), and according to BischofYs nomen- clature is named diphenyl-ary-diacipiperazine. The experiments described in the sequel show that benzylamine acts on ethylic chloracetate in an analogous manner, the first product being the ethylic salt of benzylamidoacetic acid, This compound, however, undergoes condensation so readily, that when kept for a few days, the gradual separation of dibenzyl-z-1- diacipiperazine is noticed, and in a few months the liquid becomes filled with beautiful, white needles of that compound, CsH5.C H2*NH*CHz* C 00 CzH5. Ethylic Benzylumidoacetate, C6H5*CH2*NH CHz*COOC2H,.A mixture of 20 grams (2 mols.) of benzylamine, about 30 C.C. cf absolute alcohol, and 11.4 grams (1 mol.) of ethylic chloracetate, was heated for half an hour on the water bath, in a flask connected with a reflnx condenser; no change was noticed except a, slight yellow colonring of the liquid, but on distilling off the alcohol, a semi-solid mass was obtained, consisting of benzylamine hydrochloride and an almost colourless oil. Ether was added to the cold mixture, when the oil dissolved easily, leaving the benzylamine salt as a white, crystalline powder; this, when collected by the aid of the pump and washed with ether, weighed 13 grams when dry (theory, 13.3 grams).The ether was removed from the filtrate by distillation, and the re- maining oil submitted to distillation under reduced pressure (10-20 mm.) ; it came over as a colourless liquid between 155" and 170" (bath, 210-225"), and on rectifying under the same conditions, the greater portion distilled a t 160-165" (bath, 220"). A complete distillation, however, was not possible, as the compound undergoes condensation easily a t such temperatures, the piperazine derivative separating in white needles in the delivery tube. For analysis, a freshly distilled sample was taken, and, as the numbers show, the product was not quite pure. 0.1338 gave 0.3312 COz and 0.088 HzO. C = 67-50; H = 7.30. 0.1308 ,, 0.3236 ,, ,, 0.094 ,, C = 67.47; H = 7.98.C,H,,NO, requires C = 68-39; H = 7.77 per cent. The compound is miscible in all proportions with alcohol, ether, benzene, and toluene. It dissolves easily in dilute hydrochloric acid, and platinic chloride precipitates from such a solution a yellow, semi-BENZY- AND ETHPLIC CHLORACETATE. I85 solid platinum salt. On long standing, condensation takes place, and the piperazine derivative described below gradually separates in colourless needles. Picrute, C6H5*CH2*NH* CH2-C 0 oC2H,,C6E2(N02)3*OH.-A hot al- coholic solution of 1.18 grams (1 mol.) of picric acid was mixed with 1 gram (1 mol.) of ethylic benzylamidoacetate. As nothing separated on cooling, the solution was evaporated to a small bulk, and ether added, when the picrate separated as a bright yellow, crystalline powder. After recrystallisation from warm ether, it presented the same appearance, and melted at 154".For analysis, it was dried over sulphuric acid. 0.2361 gave 28 C.C. moist nitrogen a t 20" and 722 mm. N = 12.87. C,,H,,N0,C6H,(N02)~*OH requires N = 13.27 per cenL Benzy Zamidoacetic acid (Benzylglycocol Z), CsH5*CH2*NH*CH2*C0 OZ. -A boiling aqueous solution of the copper salt described below was decomposed with hydrogen sulphide, and the precipitated copper sulphide removed by filtration. The filtrate was reduced to a small bulk on the water bath, and then to dryness over sulphuric acid in a, partial vacuum ; the pale brown, crystalline residue, on extraction with boiling absolute alcohol, lost most of its colour, but ignition on platinum foil revealed the presence of inorganic matter.By treat- ment with 96 per cent. alcohol, in which the acid gradually dissolves, the greater part of the impurity could be removed, but e-ren after repeated treatment in this manner small quantities still remained, and the numbers obtained on anslysis were too low in the carbon. The purest specimen we prepared crystalliaed from water in thin, white needles, and melted sharply a t 197-198". The results of the analyses of the following salts are, we think, sufficient proof of the identity of the acid. The specimen above mentioned, although very easily soluble in water, was insoluble in all the other general solvents. Sodium Salt, C9HloNO2Na.--An alcoholic solution of ethylic benzyl- amidoacetate, and the theoretical quantity of pure sodium hydroxide, was heated for 12 hours on the water bath, using a reflux con- denser ; on cooling, the sodium salt separated as a white, gelatinous mass, which, as filtration was found impracticable, was spread on porous plates ; after some days, the white, amorphous powder left was purified by warming it with a small quantity of absolute alcohol, and, after cooling, again spreading on porous plates.The salt was dried over sdphuric acid for analysis. 0-2075 gave 0.0805 NhSO,. It is very easily soluble in cold water, but only sparingly in hot N a = 12.56. C9H1,,N02Na requires Na = 12.30 per cent. alcohol.190 INTERACTION OF BENZYLAMINE, ETC. Copper salt, (C9HloNO2),Cu.-On pouring a hot aqueous solution of the sodium salt into a hot dilute copper sulphate solution, ft flocculent, blue precipitate was immediately formed ; this was collected by the aid of the pump, and washed with waym water.It was recrystallised from a large quantity of hot water, and formed small, dark blue prisms. An anaiysis was made of a sample dried over sulphnric acid. 0.1525 gave 0.0314 CnO. Cu = 16.43. (CgHloN02),Cu requires Cu = 16.17 per cent. Hydrochloride, C9H,,N02:HC1.-Concentrated hydrochloric acid was evaporated on the water bath with ethylic benzylaniidoacetate, a large excess of acid being present. A white, crystalline mass remained, which was washed with ether, and recrystallised from alcohol. Small, white plates were thus obtained, which melted at 214-215'. 0.2228 gave 0.1586 AgC1. The salt is very easily soluble in cold water, spatingly soluble in C1 = 17.57.CgH,,NO,,HCl requires C1 = 17-57 per cent. cold, easily, however, in hot alcohol. The white crystals x-hich are formed during the distillation of ethylic benzylamidoacetate, and which gradually separate from the same liquid on long standing, consist of this compound. It is best prepared, however, by heating the ethylic salt previously described (p. 18s) to boiling, for a considerable time, under reduced pressure ; after cooling, the product is mixed with cold alcohol, in which the piperazine derivative is but sparingly sduble. The crystals thus obtained are redissolped in hot alcohol, from which the piperazine separates in white, prismatic needles melting at 170". The substance was dried a t 100" for analysis. 0.1288 gave 0.3485 CO, and 0.0723 H20.0.1218 ,, 10.25 C.C. moist N at 18.5" and 720 mm. N = 9.14. ClsH,,N20, requires C = 73.47; H = 6.12; N = 9-52 per cent. It is insoluble in water, ether, and light petroleum, easily soluble in benzene, and toluene, sparingly in cold, and easily in hot, alcohol. I t dissolves easily in cold concentrated hydrochloric acid, and is reprc- cipitated unchanged on adding water ; no change was produced by heating it with the acid for four hours at 180". If the compound be heated to boiling, at the ordinary atmospheric pressure, i+ rapidly darkens, toluene and ammonia being formed. All attempts to obtain crystals from the dark brown residue proved futile, although by treat- C = 75-76 ; H = 6.23.CONDENSATION PRODUCTS FROM BEMZYLAMINE, ETC. 191 ment with hot concentrated hydrochloric acid a small quantity of the original substance could be isolat'ed. A molecular weight cleter- mination, by Beckmann's method, gave the following numbers. I. 11. Weight of substance taken.. .. 0.1551 gram Or3746 gram Weight of acetic acid ........ 11.38 grams 11-38 grams Observed depression ......... 0.162" 0.3995" Molecnlar weight found ...... 0.328 0.3.25 Zurich unicersify. Theory for C,,H,,*R'.02.. ......... 294

ISSN:0368-1645    

DOI:10.1039/CT8946500187

出版商:RSC    

年代:1894

数据来源: RSC